Why does Bridge blow out thumbnails when adding new images?

Question

Using Win 7 (64b) with CS6 and Bridge 5.0.2.4 x64 (CS6). I use Bridge in my workflow to catalog and tag images. I added 7000+ older photos in various file formats, Canon CR2, TIF, JPG, and PSD to my single directory containing all shots. As Bridge created thumbnails for all newly added shots, everything was looking good but then I noticed Bridge went back and lightened many of the thumbnails by ~ 1 EV. A few images were lightened by as much 2-3 EV and unusable (non-identifiable images).
When launching the non-identifiable image in Photoshop CS6 from bridge, tonal values and exposure are just fine. The images can be edited  and saved. Tried solving by deleting image from directory. then restoring it (with different filename) before restarting Bridge. In one case, it worked. The newly renamed and regenerated thumbnail was fine. In other cases, no matter what I do, the image blows out.
This is quite frustrating. Thumbnails just plain look bad. Don't have a lot more hair to pull.
Thanks for any help

Answer 1

Hmmm. That's not very convenient and it doesn't prevent my mistake. At the cost of a lot more typing, it can prevent other scripts from accidentally referencing or colliding with your variable names, but it doesn't prevent you from accidentally colliding with other people's variables if you leave off the object qualifier when you are typing.
When I saw a reference to namespaces in something I was reading, It thought it was more like real namespaces (e.g. C++ namespaces) where you define a scope where all declared or referenced variables are automatically in that namespace unless they are explicitly referenced/declared otherwise. Does JavaScript not have that capability?
It would be more like this:
namespace MyProgram {
var i = 3;
// lots of other code here
With a real namespace, "i" wouldn't be a global variable. It's real name/declaration would be MyProgram.i or MyProgram::i depending upon the syntax.
Does JavaScript not have real namespaces?
I get that I can explicitly declare everything to a scope of MyProgram to simulate some benefits of namespaces, but that's a lot of extra typing and, presumably a bunch of extra work for the interpreter since there's at least one extra dictionary lookup for every variable or function reference.
--John