Soldering On

Posted on by lhl
Posted on by lhl

ARE FILE SWAPPING NETWORKS CACHEABLE? CHARACTERIZING P2P TRAFFIC (PDF):

Our analysis of the traffic computed a 67%

byte-hit-rate which compares favorably with web caching

hit rates known to be in the range of 30% to 60%. Further, it

was shown that the disk space required for effective caching

of P2P traffic is small enough to be practical – close to

maximal caching is attained with 200 GB disk space.

Finally our analysis concludes that the byte-hit-rate

computed at our installation correlated with the traffic

volume, indicating that a higher byte-hit-rate may be

expected on links with a heavier traffic load.

Posted on by lhl

I got an email from Tristan Louis about his proposal for a full disclosure xml feed, but well, again, I have to ask, what’s the point? Do you exchange a full disclosure xml feed with people you meet at a party before talking to them? Hint: if you think this would be useful in your day to day life, you are probably hanging around with the wrong people. (or perhaps you simply need a better rl-FOAF protocol?)

(this is totally ignoring the point of why a marketer would include this feed and getting browsers to adopt interfaces for parsing and displaying them. still, makes for fun conversation I suppose)

Chris probably has done the most in-depth (and interesting) research on this (which includes a ragin cow network count, all the urls, and an interview w/ the Director of marketing for this project. (And yes, Casey is cute, but I have better things to do with my $19.95 a month)

Much Ado

Posted on by lhl

A lot of people seem to be in a huff over the Dr. Pepper thing, and since I had written most of something substantive on the boingboing discussion, I thought I’d post some of it here too.

[todo: tool to automatically link various web community profiles / postings into site]

Any, following jjg’s links, it looks like people are getting riled up over nothing.

Personally, I don’t think the Project Blogger thing will work, but even if it does, who cares? Good for them, whatever. It’s a whole different “blogging” universe. And unlike link spamming, it’s not going to affect us.

Why? For the same reason that it’s not a problem in real life. Because the whole point of blogging is forming decentralized trust network which approximate real trust networks. Granted, distributed (individualized!) reputation/indexing tools are still being formalized and developed (re: why did Google buy blogger), but it’s really just a formalization of what we’ve been doing from the get-go.

Whether people shill online maps exactly as whether people shill in real life or not. Obviously you have the choice of hanging around Amway affiliates or not (hmm, unless they’re family, I suppose). But hey, if it’s socially accepted in the crowd you hang around, and it’s useful, go for it. Vice versa, if your mores or self-dignity makes you averse to shilling, then you won’t, and you need never associate or even know that such a thing happens.

BTW, a quick search on indexes show that these sites are very low on indexes (if they exist at all!), so even in current tools they wouldn’t really affect the ecosystem much. And once the next generation of distributed tools are deployed shills won’t affect you at all (unless you really do trust them, which is your problem).

Posted on by lhl

The SANS Sendmail and Snort Vulnerability webcast just finished a little while ago. If you’re running Sendmail you’ll probably want to update soon before an exploit pops up. I decided to switch to Postfix while I was at it on my server. Was amazingly easy, took less than 10 minutes to set up.

Hah. ISS marketing made a flash infographic on it. Well, it’s Sendmail, so I guess they figure they’ll be able to use it over and over again.