Y'know, while I'm actually using this thing for the first time in a century...
Has it ever bothered anyone how amazingly inefficient current battery packs are with their geometric construction? Inside the nice monolithic battery packs you see out there for laptops, cordless tools, and the like, there's (typically) inside a large slab of soldered-together rechargeable variations on the everyday AA battery. Sure, they've got exotic chemistries, and they have tabs for soldering, but they're just a variation on a theme, and wasting so damn much space.
Nerd alert: If you think about it, actually.. say you have a battery pack of dimension (width, height, depth). There's w*h*d cubic units of volume in there, for whatever chemistry you care to fill it with. Now suppose you're constrained to filling this w*h*d with cylinder objects. Arbitrarily, I'm going to orient the length of the cells with the d dimension, set the diameter of the cells to h, and constrain dimension w to integer multiples of h. Our individual cell volumes look like (pi*(h/2)^2*d), multiplied by (w/h) cells. With a lil algebraic manipulation, you have (pi/4)*(w*h*d)... (pi/4) is ~0.785.
So, basically, you're throwing away ~21% of your total battery capacity by using this stupid construction technique. Actually more, since the individual cells have their own external packaging, thermal spacing, tabs, etc. Prismatic cells are a step in the right direction, but for fuckssake, they really need to be building *slab* cells. Yeah, I know there's engineering constraints like thermal migration, bulging, etc, but there's problems to be ironed out all the time. This is just delaying the inevitable.
PS: Extra credit - open up a 9V battery, and learn what fear is.
Cool - the powered picatinny rails idea I wrote about in 2005 is being worked on:
http://www.dtic.mil/ndia/2009infantrysmallarms/wednesdaysessionvTorbjoern.pdf
Looking back and reading my page, I'm pleased that the XM8 died, too.