Skip to main content

· One min read

I had a problem where df and du disagreed with the amount of disk usage. The cause was processes holding on to unlinked files. Running the following identified the processes:

ls -ld /proc/*/fd/* 2>&1 | fgrep '(deleted)'

I killed the processes and df is now showing the correct information.

· One min read

I just found out that dreamhost supports ipv6 (has done since March '11). Just go to manage domains in the control panel and click "Add IP" next to each of your domains.

Now if only our enterprise hosting at Verio/NTT supported ipv6!

· One min read

It seems that most browsers are focused on speeding up javascript execution now as most websites are heavy with ajax and complex js.

With Firefox releasing their pre-beta "dev" version Aurora this morning I decided to run all the current browsers through sunspider to see if any progress has been made.

Chrome-dev: 294ms
Aurora (FF dev): 324ms
Firefox 4: 348ms
IE9: 251ms
IE10 preview 1: 254ms

So it would seem that Firefox has sped up their js engine, but has a little way to go to catch up to the other browsers.

· One min read

We found that Google had indexed a site that shouldn't be indexed so I setup a robots.txt file to deny all crawlers and locked down the site with http auth. I also put in a request to have the urls removed from the index and cache.

When I did this Google returned ~2,400 results when doing a "site:www.site.com". A few days later it was returning ~54,000. Today it is returning ~133,000.

I'm not sure how Google managed to mix up "remove my site" with "index it more". Maybe this is just part of the removal process?

Update: Google is now up to 217,000 results for this site. Maybe removing your site from the index is good for SEO?

· One min read

Found a small command to run through code directories and change the CVS path, handy if you're changing CVS username or path.

find . -name 'Root' -exec perl -pi -e 's/OLD_URL/NEW_URL/' {} \;

· One min read

I had a problem where Googlebot was indexing a development site, so we locked it down using apache basic http auth. Now Googlebot was being served with a 401 when accessing the site, but because it had no stored robots.txt it was persistently trying to crawl the site.

Using the following allows anyone to access robots.txt but denies access to the rest of the site:

<Directory "/home/username/www">
AuthUserFile /home/username/.htpasswd
AuthName "Client Access"
AuthType Basic
require valid-use  

<Files "robots.txt">
AuthType Basic
satisfy any
</Files>
</Directory>

Eventually Googlebot will get the hint and stop indexing the site and we can remove existing content using webmaster tools.

· One min read

Redhat likes to run old packages with backported patches, but sometimes it's nice to have the latest version that includes new features. Unfortunately the unzip from RHEL yum can't unzip files > 4GB. I found the src rpm and compiled binaries that work with RHEL 5.6 (may work with older).

unzip-6.0-1.i386.rpm
unzip-6.0-1.x86_64.rpm

· One min read

I've been having problems recently with it overheating. It's not much fun when trying to code on the couch with a boiling hot mac on your lap.

I found a program called smcfancontrol which seems to take care of that. It's now under 50oC and coding it now more fun!

· One min read

Why is it that in this day and age it takes a day to transfer money from one bank to another? And if it's a weekend or public holiday it takes even longer. Do computers not work weekends now?

Can someone please invent a protocol or process to allow bank transfers to be instant? Surely it can't be that hard, it's just numbers/data being shifted around. I can transfer entire DVDs of information from one side of the world to the other in a short period of time, why can't a bank process things in real time?