Ubuntu 9.04 on Dell Optiplex 760

The issue was reported as Bug #348694.

I experienced the same problems reported there. When installing from liveCD, during the first stage you have to push the power switch to generate interruptions until the keyboard is on. Then you can start to strike a key. When the mouse is on, you can move mouse to keep the system awake. I am not sure if there is way to keep the mouse always busy in that way you will save time for installation.

The next day morning, you will find installation is finished, and you need to restart. Then you need to set the kernel options when the grub screen is on.

• hpet=disable helps, but does not solve the problem.
  
• acpi=off solve the problem, but you will lose a CPU.
• acpi_skip_timer_override is the best solution.
  

Google search is the biggest server-side mashup

Each search result page is a mashup of multiple pages, their snapshots, and possibly different representations. Of course, some contents are old, but some "important" pages are refreshed constantly in Google's index. Now they even add features like SearchWiki to personalize it. Maybe people will like the idea of SearchWiki more and more. In fact, we have used a personal mashup for a long time --- the Google Reader.

HP printer on Ubuntu 8.10

The lessons I learned by trial and error about my Laserjet 1020.

1. foo2zjs works sometimes, but will have problem when CUPS is updated.
   
2. HPLIP always works if you install the latest version, not the one in repository.
   

A page about the wide finder

I have published the page of wide finder, which was created for a group discussion on Erlang.

The problem of wide finder was coined by Tim Bray, and is summarized in http://www.tbray.org/ongoing/When/200x/2007/09/20/Wide-Finder

The task is to read an Apache logfile and figure out which pages have been fetched the most.

The reference Ruby code

# 11 lines

counts = {}
counts.default = 0
ARGF.each_line do |line|
if line =~ %r{GET /ongoing/When/\d\d\dx/(\d\d\d\d/\d\d/\d\d/[^ .]+) }
counts[$1] += 1
end
end
keys_by_count = counts.keys.sort { |a, b| counts[b] <=> counts[a] }
keys_by_count[0 .. 9].each do |key|
puts "#{counts[key]}: #{key}"
end

Bray's original idea was to check if Erlang is good option for this problem on multi-core platforms.

Read more ...

Be careful when you see some CFP's or invitations

Some links for this sensitive topic:

http://diehimmelistschoen.blogspot.com/

http://netdriver.blogspot.com/2009/01/from-site-httpwwwanti-plagiarismorg.html

http://delicious.com/tag/fakeconference

The reverse C10K problem

The original C10K problem studies how to provide reasonable service to 10,000 simultaneous clients or HTTP requests using a normal web server. I call the following problem the reverse C10K problem, or RC10K --- how to support 10,000 simultaneous outbound HTTP requests running on a web server. The RC10K problem can be found in scenarios like service orchestrations and server-side mashups. A server-side mashup needs to send several simultaneous HTTP requests to partner services for each inbound request.

Many approaches to improving the performance and scalability of HTTP servers can be applied to tackle the original C10K problem. However, whether these approaches can tackle the reverse C10K problem needs to be verified. My recent paper discussed the RC10K problem, and presented some experiment results got from a prototype mashup server application.

People like to try what they are not good at

From xkcd.com