In this case “modules” refers to the utility which uses TCL scripts to modify your environment (paths, environmental variables, etc). For example, if you want to change compilers, you can use modules to change all the environmental variables/library paths/etc that need to change, and if you want to change back, unload that module and it changes them back.

Now this is really nice, but there are some things I’d like to see changed:

1) In our implementation, the initialization script “.modulerc” is read every time module is invoked. So, if you have .modulerc loading a certain module, and decide that you don’t want it, and unload it. Now you list the modules to verify that it works, the module is reloaded because you called the module command to do the list. There are other commands which end up modifying your .modulerc file, but those are not well supported (by us anyhow). Plus, `module unload something; module list` would be very counterintuitive… it makes no sense.

We just don’t use the .modulerc files, and instead have shell files which we can source to change the environment, but a) .modulerc causes problems for unknowing users b) a .modulerc file would be a good thing to have, sourcing shell scripts is more or less a workaround.

2) This is more of a wish-list item: our library modules take a look at which programming environment module you have, and use the appropriate libraries. Default libraries (Libsci, fftw, totalview) are automatically loaded/unloaded by the Programming environment modules, but other libraries have to be manually reloaded if you change libraries. It would be nice if there were flags so that all of the “library” modules were automatically reloaded when a “compiler” module was loaded.

3) Not that this will change, but the name is unfortunate: ‘module’ means so many things (or perhaps the same thing in so many contexts), that if you google for ‘module’ looking for this program, the S/N is terrible.

At the moment, battery technology is not ready.  Right now, the battery of choice seems to be Lithium Ion.  Let’s look at the Tesla Roadster: Wikipedia cites that they have 6831 Li-ion cells, and a range of 255 miles.  I assume this range is based on all brand new batteries?  Let’s say it is for argument’s sake.

Again, Wikipedia says Li-ion’s lose 20% of their capacity a year if kept fully charged at 25°C.  Let’s pretend that is 20% of the usable portion of the battery’s power.. I have no proof, but but that seems like the best case scenario.  If so, you would lose about 20% of your range a year.  After a while, you would need to replace (some of) your batteries.  The regime that would make the most sense would be to replace a certain percent of your batteries every year (or twice a year, whatever).  Let’s say you replaced 20% of the cells every year, replacing all of them in the course of five years.  Then, you would end up with 67% of your total capacity immediately after replacing the batteries, and 54% immediately before.  Perhaps you’d want to replace every six months instead, but the point is that you wuld have about 60% of the maximum range.  If that is 255 miles, your real range would be 153 miles.

Now you could save on batteries if you didn’t need such a big range, or you could increase the range by buying more batteries.  Ha, maybe the poor could get used batteries at steep discounts so we don’t have to recycle still-decent batteries.  Anyway, with the Tesla, that would constitute almost 1400 Li-ion cells per car per year.  If they were 6 cell laptop batteries, 227.7

So let’s say a significant portion of Americans switched, that would be on the order of 10 billion laptop batteries a year.  Infrastructure can be built, but think of how this would affect the price of Lithium… Clearly, we would be recycling these, not throwing them away, but how good can that be?  Maybe it can be great, but it seems like a huge challenge to me.  Batteries work well in the small scale, but I don’t believe that they scale up so well.

And the same thing goes for fuel cells right now: we need platinum for the electrodes, but there simply isn’t that much platinum.  You can make one fuel cell, you can’t make hundreds of millions big enough for cars with the current technology.  So they are working on other types of electrodes, I have heard a lot of research about that sort of thing, but I don’t think any of it is ready quite yet.  Soon hopefully.

There are mechanical storage mechanisms such as pneumatics and flywheels.  These appear to be viable for hybrids at the moment, however, they strike me as dangerous.  Maybe not though, we do carry gas around..

I think the most promising route is metal-air fuel cells.  Here, the fuel would simply be aluminum or zinc, both of which are ridiculously common.  You would buy pellets of metal and return metal-oxide paste.  It would be extremely safe.  In fact, we might have to stop using pennies if we went to that design since they are mostly zinc, and you probably get more by changing a $20 into pennies than buying $20 worth of zinc.

Of course, that seems to be a long way away too: I do not believe the design is mature enough to work yet.  But if I ran an energy company, that’s where my money would be.