techno.blog("Dion")

JDO 2 has been voted through.

Not dead after all (in the JCP eyes at least ;)

Even Hani voted yes (with one No Vote from JBoss)

I am always frustrated when I see “input” being the metric of choice at companies, and in projects.

Paul Graham talked about this in his talk at OSCON (a lot of good stuff there).

An obvious example that I saw in a different sphere, was seeing a great football (soccer) referee having to quit the big time because he was past the age restriction.

So, no matter if his performance in games are top in the league. Or, if his eye-sight is tested and it is fine. Or, if his fitness is fine. It doesn’t matter. Turn a certain age and YOU ARE OUT.

For gods sake, kick out the refs that are not performing, not the experienced ones that are.

I see this kind of thinking in corporate world all the time.

I was talking with the head technical honcho of one of the large Massive Multiplayer games.

I worry a lot about scalable performance, but I had to laugh when I compare the issues that I have had in the enterprise, to his.

Imagine having to rewrite firmware for routers and such because it is the bottleneck. Firmware. Ouch.

He has a tough problem because his world is all about tiny packets, and a huge amount of them. Needing low latency in that environment is a tough problem.

Now I can feel good about the problems that I am having to solve at the moment :)

*walks off to recompile ciscos os*

Yahoo! Photos is magnitudes larger than Flickr, so why would Yahoo! buy them? Del.icio.us is surprisingly small compared to the mindshare too.

Does Yahoo! see growth? I don’t think that they really care.

Both can give them a lot to learn about the web, which can help the search business get more relevant. A holy grail of search is to allow me to get results about Apache Ant vs. the insects when I search for “ant”. Contextual search. Some talk to the semantic web as a solution.

What Yahoo! now has with their properties are dedicated users who are doing the semantic tagging for them. The community is doing the work, and humans and the wisdom of crowds are better at it than the best AI right now.

A danger is that the del.icio.us user base is probably skewed to a particular set of stereotypes (tech savvy etc), but that may not matter much for them right now.

Yahoo! got communities for pretty cheap, and they don’t even need them to grow to be useful.

I use a bunch of editing tools. OmniOutliner, TextMate, IntelliJ IDEA, vi, the editor that lives in my email client, etc.

One of the great things about TextMate is that it groks the power of the shell, and of the platform.

This is shown throughout the product, from being able to type a command and have it kick out, run it, and put the results back in the document, to the plugin system.

I have written a bunch of plugins in TextMate to make my life easier. Why haven’t I done this in IntellIJ or Eclipse? The thought of doing those is daunting. Time consuming.

With TextMate I can write a plugin in any language that I want because the contract is so loose. It will shell out, and I can run ruby, perl, lisp, or whatever I feel like for the given problem.

This is a truly killer feature.

I had a fun three visits to the Apple Store in SF today. The line was long as always, and I wished that I hadn’t picked up the Plantronics .85 headset.

The reasons that I went for it were:

• USB piece is separate, so I can plug it into my PCs and Mac
• Folds up small, yet fits ok

Unfortunately, it just didn’t work. The sound quality on the mic was painfully bad. Unusable.

Back to the Logitech.

It still bugs me that the Mac needs a powered mic. I mean come on. Let me plugin a normal damn mic and have it work!

I don’t know about you, but all I have felt is pain when I import and export OPML between various RSS tools.

It has been such a nightmare that I have had to work out the quirks in the given reader, and write scripts that munge from one to another.

Well, hopefully that will be fixed, at least if OPML 2.0 has anything to do with it.

Or, will we have AtomOPML 0.3b?

Back in the day, I was the first person to know, and care about version 2.6.1 of FooBar, the open source framework that does everything that you need.

For one, as editor-in-chief of TheServerSide, it was my job to be on top of things.

For two, I actually cared. The playing field was fun, there was a lot of innovation. It was a brave new world.

Fast forward to 2006, and I am the anti-framework releaser. If I never see “YetAnotherMVC 1.2 Released” announcement on TSS and others, it will be too soon.

Isn’t it just so BORING? Partially, I think I hit the end of the road on keeping up with the Jones’. At the end of the day I want to get my job done. I want to produce business functionality for my clients. ROI and all that garbage.

Now, I am living part of my life in the land of the UI. What has excited me about this land, is that when I develop a great Web UI for a client I can:

- Show my wife, and she can see that it is cool. Telling her (totally non-technical thank god) that I refactored my 3-tier backend using ALL of the GoF patterns isn’t a turn on. Neither is showing her a web page, but at least she can get it.
- Same for the client: Watching clients say “wow” is worth it. The backend is very important, but the front end is big too.

So, I am officially bored of all of the server side framework crap. I do listen for some things. Every Dojo/Prototype/Scriptaculous release makes me look, as so much has been added, because they are so young.

Aren’t you kinda bored too?