I think I figured some stuff out - slashdot is indeed performing an HTTP connection back and trying to retrieve "ok.txt", but it's infrequent and has been going on for some time. The connections are all from slashdot.org 216.34.181.45 and started about 10 months ago. If I check slashdot itself for the file (http://slashdot.org/ok.txt) it exists, so a possibility is that slashdot is testing to see if my connection is coming from an anonymous web proxy, though I don't know why they'd care. The file returns "ok", so I'll try putting that in and see what happens. I'm surprised that nobody else is noticing this. I'm certain it's not just me.

The slowdown seems to be the result of a really massive pile of CSS/javascript stuff coinciding with a spam attack and my wimpy DSL line. I don't know if the CSS craziness is new, but I'm blocking it all, and slashdot works fine and is much cleaner and faster for me now.

I know this is completely offtopic. Apologies. Slashdot doesn't have a static user support thread that I'm aware of, or I'd have asked in it instead.

Something seems to have changed with slashdot recently. It started connecting back to me as if I were a web server to get a file called "ok.txt", so I started throwing some random stuff in that file to send back. Then I noticed a connection being made from my browser to c.fsdn.com to download 400K of something or other. This makes slashdot takes so long to come up that it's pretty much unusable. Is this happening to anyone else?

After wading through quite a bit of United States Code (in particular, subfunction 42 1395y and functions called by its caller) trying to grok how the Virtual Justice Machine might interpret the situation, I discovered 1395b, which seems to say that the Medicare law cannot be construed to mean that someone can be denied health insurance because of it. I pointed the insurance company at USC 42 1395b and a week later, I'm insured.

It looks like Medicare was given a mandate via some changes to 1395y that went into effect on Jan 1, 2009 to come up with a reporting mechanism that any group health insurer carrying someone who was also eligible for Medicare would have to use. However, they made the mistake of implementing it with a design that requires insurers to transmit subscriber SSNs to check their Medicare eligibility status, and then when the insurers starting asking what they're supposed to do about all the people for whom they do not have SSNs, Medicare hinted that they should just drop them. As someone who ran into problems buying health insurance without revealing my SSN because of this, I'm concerned about the idea of having to use an SSN to buy anything.