Misc. Ramblings

Filtering spam probably will never be 100 percent effective. Why? Because it is possible to create an e-mail that does not use any of the standard spam techniques but still get its message across.

For example, I got one last week that scored 0.0 in SpamAssasin. That is, as far as SpamAssasin could determine, there was nothing to indicate it was spam.

Even the text seemed to say that the sender had found a broken link to their site on one of my pages and would I like trade links with them by just clicking on the handy HTML link at the top of the e-mail?

Of course, I had no such link to their site and clicking on their handy button would just confirm that they had reached a live e-mail account that they could then sell to other spammers.

So, what to do? I don't know. All I can say is the most effective deterrent would be one based on economics. That is, one that makes it cost more to send spam than the income they get back from it.

When someone figures how to efficiently do that, they will be crowned king of the world.

Opera 7.2 is Gold. The latest version of the Opera browser went out the door recently (get it here). While I seem to be using Mozilla Firebird as my browser of choice right now, Opera still has it uses. For example, when I think I need a hardened, bullet-proof browser to wade into enemy territory, it's Opera I turn to lock and load (mixing metaphors). Yes, Firebird may be faster, but how secure is something that hasn't even reached the 1.0 release yet? Especially one that specifically states it should not be used for mission critical tasks!

So when the going gets tough, get Opera.

To the Moon. Speaking of going places, the European Space Agency is set to launch their first probe to the moon. The SMART-1 (Small Missions for Advanced Research in Technology) propulsion system includes a main ion-drive engine powered from a solar cell array. See the story here.

One Correction from yesterday's post on the Apple array. I did not realize that Virginia Tech was looking only at 64-bit processors (Thanks to Ken Scott for the information). As such, the cost of the G5 is probably (although I have not checked this) lower than the alternatives they looked at.

Of course, as Sjon pointed out to me, they could have also looked at AMD. In addition, as I understand it, the point of a massively parallel array is to use the power of as many nodes as possible. Regardless of whether they are 64 or 32-bits.

For example the Maui High Performance Center Linux SuperCluster uses 260 nodes with a pair of Pentium IIIs per node. At the time of its power-up, this cluster was the worlds most powerful array there was and, for all I know, still is.

It seems to me that if you spent less money on getting 64-bit processors and instead got a bunch more of cheaper 32-bit ones you would reach a point of higher total power. On the other hand, what do I know? I could be wrong and Apple will reign supreme in the 64-bit super computer world. Not.

Aloha!