Five onions, half a stick of butter. Slice the onions thin, and sauté on medium heat for a bloody long time — about forty-five minutes, making sure it doesn’t stick. The onions will be distinctly brown after treatment such as this.

Add a quart of water, three-quarters of a cup of pearled barley, half a dried serrano pepper, five allspice, two tablespoons of soy sauce, two drops of liquid smoke (quite optional but delicious), a pinch of rosemary, and a teaspoon of salt.

Simmer until the barley is tender, half an hour or so.

Serve with a grating of goat gouda or gruyère cheese, broiled if you can get it to float that long.

Looks like OpenWRT turns up the transmit power to 255 when it starts up on the Linksys WRT54G. Not good…

WiFi IEEE 802.11b/g ESSID:“theinternetco.net” Nickname:“acrux” Mode:Managed Frequency:2.462 GHz Access Point: 00:0F:66:E3:FC:A8 Bit Rate:24 Mb/s Tx-Power=31 dBm Sensitivity=20/200 Retry min limit:8 RTS thr:2347 B Fragment thr:2346 B Encryption key:off Link Quality:255/0 Signal level:-41 dBm Noise level:-7 dBm Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0 Tx excessive retries:0 Invalid misc:0 Missed beacon:0

Good strong signal … outpaced several orders of magnitude by noise.

After a wl -i eth1 txpwr 40…

WiFi IEEE 802.11b/g ESSID:“theinternetco.net” Nickname:“acrux” Mode:Managed Frequency:2.462 GHz Access Point: 00:0F:66:E3:FC:A8 Bit Rate:54 Mb/s Tx-Power=31 dBm Sensitivity=20/200 Retry min limit:8 RTS thr:2347 B Fragment thr:2346 B Encryption key:off Link Quality:16/0 Signal level:-55 dBm Noise level:-126 dBm Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0 Tx excessive retries:0 Invalid misc:0 Missed beacon:0

Nice and clean!

The ZoomQuilt is among the coolest art projects I’ve ever seen. Thanks Erin.

Making yet another functional language, re-inventing LISP, is like making square wheels because thirteen-sided ones bother you.

I love IPv6. It’s this easy to set up a new host:

root@centauri:~# insmod ipv6 Using /lib/modules/2.4.20/ipv6.o root@centauri:~# ip addr 4: br0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue link/ether 00:0f:66:e3:fc:a6 brd ff:ff:ff:ff:ff:ff inet 172.16.17.245/24 brd 172.16.17.255 scope global br0 inet6 fe80::20f:66ff:fee3:fca6/10 scope link inet6 2001:470:1f01:301:20f:66ff:fee3:fca6/64 scope global dynamic valid_lft 2591997sec preferred_lft 604797sec root@centauri:~# ping6 nbtsc.org PING nbtsc.org (2001:470:1f01:301:208:c7ff:feca:2758): 56 data bytes 64 bytes from 2001:470:1f01:301:208:c7ff:feca:2758: icmp6_seq=0 ttl=64 time=19.3 ms

Add multicast DNS or anycast DNS, get the IANA to register a well-known DNS anycast address, and getting on the internet will require zero configuration, and not require ugly systems like DHCP. There simply won’t be anything to configure.

Thank RedHat for RPM.

My computer did some serious evil to itself today. I fixed it in a couple hours, thanks to RPM’s database:

rpm -qa > rpms then for I incat rpms; do echo $I:; rpm -V $I; done | tee log

Those two commands give a list of what’s damaged and how, and based on that, makes it darn easy to restore stuff.

User data is something else entirely, of course, but there’s a reason I keep that on another drive and make backups…

Compare this to the nearly identical but much more laughable situation I dealt with this afternoon on a customer’s computer. Windows, having no central understandable database of what’s been messed up, nor any way to restore just one corrupt file (unless it’s a DLL or other immutable system file). I spent three hours tracking down the fact that the system registry hive file was damaged, as was the backup, then had to re-install all of windows to make it function again. The only reason it didn’t take ten hours to fix is that this particular customer has three important files: his inbox, his address book and his quicken data store. Nothing else mattered. A re-install took as long as my fixing my own computer did in total, and the damage to mine was far, far worse.

Lessig asks questions, Powell evades with the usual.

I must say that Instiki isn’t the most efficient wiki implementation I’ve ever seen.

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 32296 aredride 16 0 869m 543m 8224 S 0.0 53.7 14:31.37 ruby 32354 aredride 16 0 869m 543m 8224 S 0.0 53.7 0:00.00 ruby 32355 aredride 15 0 869m 543m 8224 S 0.0 53.7 0:00.07 ruby

This was in our access log. Nothing removed to protect the guilty, either.

217.8.227.181 - - [26/Nov/2004:02:54:59 -0700] “GET /wiki/NBTSWikiWiki?edit HTTP/1.1” 200 7843 “http://www.google.ru/search?q=wiki++inurl:edit&num=20&hl=ru&lr=&start=120&sa=N” “Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; dial; .NET CLR 1.1.4322)”

That a GET for the edit page directly, referred by google.ru.

217.8.227.181 - - [26/Nov/2004:02:55:02 -0700] “GET /style.css HTTP/1.1” 200 984 “http://community.nbtsc.org/wiki/NBTSWikiWiki?edit” “Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; dial; .NET CLR 1.1.4322)”

217.8.227.181 - - [26/Nov/2004:02:55:04 -0700] “GET /wiki.css HTTP/1.1” 200 707 “http://community.nbtsc.org/wiki/NBTSWikiWiki?edit” “Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; dial; .NET CLR 1.1.4322)”

Getting the style-sheet. Weird, for a robot, but not unheard of. By the user-agent tag, it might be an automated Internet Explorer process. Six seconds.

217.8.227.181 - - [26/Nov/2004:02:55:34 -0700] “POST /wiki/NBTSWikiWiki HTTP/1.1” 302 - “http://community.nbtsc.org/wiki/NBTSWikiWiki?edit” “Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; dial; .NET CLR 1.1.4322)”

Standard POST, just like all edits. I did not log what fields were filled in, but it might be interesting to see. Thirty seconds. Done by hand?

217.8.227.181 - - [26/Nov/2004:02:55:50 -0700] “GET /wiki/NBTSWikiWiki;1.255 HTTP/1.1” 200 65875 “http://community.nbtsc.org/wiki/NBTSWikiWiki?edit” “Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; dial; .NET CLR 1.1.4322)”

A GET on the updated page. Probably just because IE does it, not because they’re checking their work. Twenty-five seconds. Or maybe they just have a slow (or distant, they’re coming from Siberia) connection.

Whois says they’re from Siberia, anyway:

inetnum: 217.8.224.0 - 217.8.235.255 netname: SCS-900 descr: Siberian Cellular Systems - 900 descr: GSM provider in Novosibirsk country: RU admin-c: SY27-RIPE tech-c: SY27-RIPE status: ASSIGNED PA notify: hostmaster@scs-900.ru mnt-by: SCS-MNT changed: ip-dbm@ripn.net 20021021 source: RIPE

SORBS says that perhaps that’s a façade:

Address and Port: 217.8.227.181 Record Created: Mon Sep 20 06:39:07 2004 GMT Record Updated: Mon Sep 20 06:39:07 2004 GMT Additional Information: Likely Trojaned Machine, host running Korgo3 trojan Currently active and flagged to be published in DNS

In the process of writing a LALR(1) parser for the NBTSWikiWiki, and then seeing the front page get spammed a second time, I wonder now if stricter parsing would have avoided the problem. If pages couldn’t contain invalid markup constructs, and you just rejected the edit…. Then that last spam would have failed since NBTSWikiWiki’s syntax is a bit different than the spammers were suspecting.

It’s probably a bad idea, but it’s a thought. In a more geeky use of a wiki, it might be very smart.

This is totally stupid, but it made me think of Fuzzhead.

One of the most bizarre things to track down in a while, but this is what happened. We had a standard left-column navigation. An unrelated block was set with margin-top: -10px;. By changing that to position: relative; top: -10px;, things started working. Talk about a bizarre bug.

Ego surfing is fun. There’s 13,100 references to me out there.

Oddly addictive.

P.S. I’ll find my frog.