Killing Code Names

August 16th, 2006

As part of the process to open source Java, Sun have switched from using codenames for Java releases to just using the actual name. Thank goodness for that. I'm not sure I know of anything more frustrating than attempting to have a conversation with someone using code names instead of actual versions. Was Dolphin before or after Mustang? Which Tiger do you mean, Mac OS Tiger, Java Tiger or Tiger Direct? How hard is it to just pick a version number and use it? Does marketing really need to mess around with that number at all, and even if they do, do they really have to wait until the very end of development to pick a number?

Ephox has for some reason decided that code names are absolutely essential to have and that we couldn't possibly refer to projects by the version numbers that engineering and product management have been referring to them as for the past year or more. Nope, got to pick a new name and then remind everyone what version number that maps to every time you use it.

I don't complain much about Ephox and it's probably a good sign that all I'm complaining about is the use of code names - apart from causing confusion at the start of every meeting it's really not that big a deal.

On the plus side, sanity reigns in the engineering department - we just focus on the story cards and get stuff done, the rest of the company can worry about what cards make up the Harrison release and which are part of Moreton. Bonus points to any Ephox employee that knows what the Buzz Lightyear release was. Of course, anytime I have to pick a project code name you can bet it'll be really memorable…

Improving The HTML Validator

August 16th, 2006

Sam Ruby - Time for a new HTML Validator?:

If people feel that HTML 5 deserves a better validator, I’m willing to invest some time into coding. Are there others interesting in contributing to the coding, the writing up of test cases, or the authoring of documentation?

 Is there a reason we need a new validator instead of just improving the existing one? The source code is available in CVS and there's some brief instructions on how to contribute. Admittedly it doesn't look like the easiest project on earth to jump into but it looks possible to do so. Is noone around to accept and apply patches?

There does seem to be quite a long bug list that people could sink their teeth into if they wanted a better validator.

So, is there a reason we need to start again from scratch that I'm missing?