The Dilbert Blog: Aging Brains:
To compensate for my inevitable mental decline I am already doing triage on entire categories of my memory. Anything I don’t need will be purged to make room for new stuff. I already got rid of the category I call “who wore what.” If I see you in the gym wearing a full chicken outfit I will remember that as “saw you working out.”
Man, I can so relate to that.
So I am sick today...my body is fighting a war with some virus that has been floating around the office this week. My mental capacity is fully half of normal, or at least that's what it feels like. But it's Friday so I guess that nobody has noticed yet. ;)
Highly recommend the Airborne for when you start to feel the slightest bit sick. It really does work, has kept me from getting ill many times over the past 4-5 years. I don't get Flu shots, don't believe in them. I'd rather let my body build up the immunities naturally. I get a little bit sick every year...back when I took the Flu shot, I would get really sick every other year when the Flu shot was for the wrong strain and I had no immunity built up.
Speaking of the Flu, this is a great present for any medical/nurse types.
All my geek friends know that I am a graph addict. I manage about 100 Linux servers that have every single mildly interesting property graphed 6 ways from Tuesday. I'd graph how often I go to the can if it wasn't such a manual process to record it. (FOR NOW!!) Personally, I rarely use MRTG anymore and have ditched it in favor of RRDTool...except for network gear. MRTG was designed for that and is still a good application in that respect.
Anyway, as part of an ongoing series (har) on core UNIX tools (sed, awk, grep etc) I present a small shell script that will smooth out spikes on your MRTG graphs. Run it from the directory that contains your MRTG log files. To edit the threshold cap, change 1000000 to whatever you need.
#! /bin/sh
LIST=`ls *.log`
mkdir -p old/
cp *.log old/
for i in $LIST
do awk '!(NF==5 && ($4+$5) > 1000000)' old/$i > $i
done
Download here: http://www.mybrainhurts.com/software/smooth_mrtg.txt
P.S. Thanks to Jason for the del.icio.us link that inspired me to be more interesting by blogging daily.
My La Fonera is neato. But there were a few critical details left out of the install docs for us Mac Users.
Everything more advanced than 'plug it in' references the User Manual on the CD. The CD is useless. It's got no files on it other than the Quick Install Guide (QIG.pdf) that you already have in printed form. To find the user manual, go here: FON: and click on: "Activate it when it arrives. More" It will launch a flash app that lets you download the user manual. Or you can click here: http://www.fon.com/manuals/M2100_1_01.pdf Default WPA settings don't work on a MacBook. You must log onto the router's admin console and change this. Plug the router into your mac via the included ethernet cable. Surf to http://192.168.10.1 in your web browser. Log in. The user is 'root' and the password is the WPA key for the private SSID/MySpaces. If you just opened it, it should be the S/N on the bottom of the device. Now clik on 'Private WiFi' and change the Authentication from 'WPA/WPA2-Mixed' to 'WPA2'.
Your MacBook will now connect. I assume this is not just a MacBook thing, but a Mac thing. Why wouldn't they document this? It's a huge pain for a Mac user trying to figure it out. Hopefully google searches will find this and other Mac users will learn from my efforts.
P.S. Works great now!
List of good things that happened to me today:
Overall, a nice Friday. :)
I made some mods to all my zone files just now (changed name servers) and needed to update all the serials as well. The name server change was easy enough:
# sed -i 's/from/to/g' db.*
But changing the serial is a little tricker. A serial number is in the form YYYYMMDDNN where NN is the update number for that day. So today would be 2006111100. This is the forumla:
# sed -i 's/200[0-9]\{7\}/2006111104/g' db.*
This says to replace any text starting with 200 and followed by 7 digits with 2006111104 . I should add this to the sed one liners page: http://sed.sourceforge.net/sed1line.txt
BTW, my high school web page is now on my servers.
I only now learned about this. But that's because I guess I am old now, I find out about the latest catchphrase after they're already tired. Still, I laughed my ass off when I was reading that and came across this:
IMG_4953.JPG. Tracy
and I made some pretty cool pumpkins this
halloween.
I made a time lapse of our pumpkin creators and put it up on vMix:
For the record, my pumpkin was titled "Angry Robot" and Tracy's was titled "Makes Nathan Sneeze".
Lots
of A-Holes cutting me off...pretty sure that the reason the traffic was
insane getting out of Del Mar was because everybody was rushing home to
vote. I voted! Line was long, took about 40 minutes to vote.
Problem: Most modern Digital Cameras do not actually rotate an image based on the orientation, they just set a flag in the EXIF headers. Most software (including iPhoto and Preview on the Mac) will display the image properly by rotating it on load. But uploading to online based services means you have an image with the wrong orientation. The ones that plauge me are Flickr, Wikipedia, Gallery2 and vMix Slide.
jhead + jpegtran. jhead will read the EXIF headers, detect the oritentation and then call jpegtran to do a lossless rotation on the image, saving the proper EXIF headers back to the file. What does that mean? It means you will have a file that is properly oriented for display on your editing software and for uploading to online services.
jhead: http://www.sentex.net/~mwandel/jhead/
jpegtran: http://sylvana.net/jpegcrop/
There are jhead binaries for linux, mac, windows. To install jpegtran on OS X I used the Fink software package. On Linux I used yum.
To automatically detect and rotate to the proper orientation, run this on the command line (OS X Terminal):
# jhead -autorot *
Easy and fast. I ran this on all the images I have taken with my new digicams, it only took a few minutes to run thru a few thousand photos.