We aren't going to build another Shuttle. The people and tools that built them were dispersed long ago. The Columbia design was 25 years old, and a compromise at the time.

An engineer wouldn't even sit down to spec out a home PC without asking 'What's it for?'. Certainly we need to clearly define NASA's continuing goals in space, before putting money and lives at stake with a replacement design. Krauthammer and Clarence Page both make the point in their columns, one of the few days we'll see them agree. Glenn Reynolds is there as well, and surveys commentary around the blogosphere. I personally like this post at Voyage to Arcturus that catalogs some of the reasonable goals.

I'll pass on commenting on any specific design ideas, not being an aerospace engineer. However, I will make one specific point, drawn from the land of Moore's Law: 'Reuse' of a space system could be a counterproductive goal. It may be the best way in some cases, but there's a lot to be said for learning from something, using it up, and building a new one better and cheaper. That's what we do here in the Valley - it's called a learning curve. Ignoring that lesson left seven brave people up in an obsolete vehicle, locked in by a system that couldn't improve.

Update: Jay Manifold has the numbers. After building and launching 13 units, the Saturn program was substantially cheaper per unit weight lifted, in inflation adjusted dollars, than the shuttle after over 100 launches. This is at least highly suggestive of where the most useful learning curve is to be found.
4:56:32 PM    

AOL subs drop for the first time, amid indications the company will no longer pursue marginal subscribers who 'freebie surf.' Can you say 'saturation'?
4:34:14 PM    

Biomimesis is borrowing design elements from nature to incorporate in machines. Robots like AIBO have an external similarity to natural creatures, but little similarity in their working. Many research projects, some aimed at miniaturization, are more direct in their inspiration. Once again, DARPA and other military funders are behind much of this. Some interesting examples:

• Case-Western's Biologically inspired robots group, likes cockroaches and crickets.
• Berkeley is getting microbots below the gram weight threshold, and has prototyped components of inch wide automata based on flying insects.
• Stanford also like bugs, and is working on manufacturing as well as control methods.
• Northeast University is taking its inspiration from lobsters (good New Englanders there) to create bots designed for underwater minesweeping.
• Watch the robot videos at U. Mich. So bug-like it's creepy.

3:01:30 PM    

Ballmer's been going around saying that open source is a threat to life as we know it and certainly the U.S. economy for quite a while. Now Microsoft's SEC filings specifically acknowledge that it's Microsoft's ox that is being gored, particularly its ability to extract revenue from the server sector.

Let's remember that it's not just the guerilla fighters of open source doing this, it's also big guns like IBM that can stand the losses if some systems elements are commoditized. Further remember that it's open standards between clients and servers, and among servers, that make this beneficial competition possible (should you be tempted to consume any .NET Kool-aid).
1:10:55 PM    

Salam in Baghdad reports among other things that last week senior Iraqi officials who picked up their phones heard a voice in Arabic suggesting they shouldn't attack civilians, rather than a dial tone. Switch hack, anyone? Then this morning Colin Powell plays conversations captured from the Iraqi military communications system. Destroying the enemy's confidence in their communications systems is next best to destroying the system itself. Good job, guys!
1:06:33 PM