How to stop popoing out the IE when AIM starts up.

This is a problem that I was suffering for sevaral months.  Anyway i found a way to do so. Not that its hard to do but I guess people can’t find it.

Disable AIM Today or AIM Dashboard from popping up in Internet Explorer

• Open Aim 6.5
• Click Edit
• Click Settings
• Click Sign In / Sign Out
• Click None
• Click Save

Hope this helps

Information retrieval (IR) Systems: Enterprise and Web Search

I heard many people now tends to talk about a new era of computer and information technology called Information age (Some may call it Digital or Wireless age according to his preferences). So, It’s clear that in the

present world ‘information’ (digitally represented) becomes the master and we all relaying on it.

When it comes to information, gathering and storing information is almost done up to the perfection owing to modern development of computer science. But the ‘information retrieval’ is still a challenge to the ICT world.

The area of computer science dealing with searching and processing collections of multimedia documents is known as Information Retrieval (IR). The IR systems solve the effective storing and querying of textual (or multimedia) documents.

The systems can be used for web pages searching in the Internet or a specified collection of documents are often referring to as ‘Search Engines’.

More precisely “Googling” is more or less the most popular way of information retrieval mechanism from the world’s largest collection of multimedia and textual document collection; the World Wide Web. Apart from Google, there are several information retrieval systems yet to become famous like Yahoo, Live Search etc.

Web Search Engine are quite popular but the modern business models relying on much more customizable IR system that you can plug various different collections of data to the IR system. For example a company may want to use the web, a pre-defined set of documents and some other third part IR retrieval system to be used as the company data repository to the company IR system. This is where the ‘Enterprise Search Engines’ come in to play.

Currently world’s most famous and reputed Enterprise Search Platform is the FAST Search and Transfer (now owns by Microsoft Cooperation) which is able to beat the giant’s (Google’s) Enterprise Search Platform. By the way FAST do support crawling through the internet but as a web search engine FAST is no way near Google.

So, its clear that both Google and FAST are going on two different path while both of them are the predominant leaders in each streams (web & enterprise search).

Information Retrieval use various different technologies to store and retrieve information and web search and enterprise search uses different strategies in their own methodology.

GNU Make parallelization – GNU make with –j Option

The Software Professionals often have deal with massive code bases which comprises of at least 10000 source file or more. If you want build the code base to binaries (or make with GNU auto tools) it would take at lease 10 hrs to build such a code base. (For me it took nearly 8 hours to build the whole source tree on a 8-core CPU server).

And many build processes run for hours with build managers commonly typing ‘make’ or some other third party build tool command and going home for the night (God sake, if the build break in between). GNU Make’s solution to this problem is parallel execution: a simple command-line option causes GNU Make to run jobs in parallel using the dependency in the Makefile to run in the correct order. For that you can use –j option to specify the number of parallel executions that you required.

-j N or –jobs N

To start GNU Make in parallel mode you can to specify either the -j or –jobs option on the command-line
The argument to the option is the maximum number of processes that GNU Make will run in parallel.

For example, typing make –j 4 will allow GNU Make to run up to 4 subprocesses in parallel. This would give a theoretical maximum speed up of 4x cutting build time by a quarter. The theoretical time is,
however, severely limited by restrictions in the Makefile.

Follow these steps before you run in to GNU make parallel build
process.

Step 1:

Verify the number of CPU cores in your build server/machine. You can use following command and count the number of CPU cores (0 – n). (say n)

less /proc/cpuinfo

Step 2:

Give the –j option with the number of CPU cores available (number
of jobs must be less than or equal to the number of CPU cores). For example you
can give,

make –j 8 <other parameters and options>

Caution:

Be careful, if you specify the number of jobs higher than
the number of CPU core, then you may have to build your code base more or less
for a weeks owing to context switching overhead.

I hope to discuss the benefits and fit falls of this method in near future.

Gmail chat not working in Opera 9.5 – Review

My earlier post on making GMail chat working on Opera 9.5 was a issue that was discussed by most of the Opera fans (i.e. I guess they are Intelligent web users).

Opera 9.51 fixed most of the issues that were sustained even in Opera’s revolutionary version 9.5. The version 9.51 basically fix all the errors on Gmail chat and you no longer have to use the specific URL that I was mentioning in my previous post. There were sevaral positive and negative comments on that post but its just make things work with opera9.5 version. Since now you no longer need to use that one, its not a big issue right now.

Although the Gmail chat works properly in Opera 9.51version, there is an issue with sudden on and offs of Gmail chat in Opera 9.51. Hopefully Opera will fix that one pertty soon.

The main reson behind these issues is that Google apps are not tested with Opera not because Opera is an obselete web browser but google never wants to do so, because they are really backing up Firefox.

SCP in 60 seconds

‘Secure Copy’ is a common but a tedious command to remember even for Linux experts. So, I’ll try to present you a clean and complete how to guide on SCP.

scp stands for secure cp (copy), which means that you can copy files across an ssh connection that will be encrypted, and therefore secured. You can this way copy files from or to a remote server, you can even copy files from one remote server to another remote server, without passing through your PC.

Putting a single file

• devcmb1> scp mydata.dat dev77.fastasa.no:NewData.txt

copy file “mydata.dat” from your current computer/directory to your home directory on dev77.fastasa.no, And re-naming the copy on newton “NewData.txt”.

• devcmb1> scp mydata.dat dev77.fastasa.no:

Same as the previous one but the file is copied as it is (without renaming).

• devcmb1> scp mydata.py dev77.fastasa.no:/se/etc/src/ linguistics.py

copy file to a specific directory on dev77.fastasa.no, giving the file a new name there.

• ktest@tframe. fastasa.no>

scp mydata.dat kasundev@dev77.fastasa.no:new.dat

copy file from your account on tframe to your home directory on dev77, where your username is different (kasundev)

Getting a single file

• devcmb1> scp dev77.fastasa.no:mydata.dat ./mydata.dat

copies file “mydata.dat” from the your home directory on dev77 to your current working directory on devcmb1

• devcmb> scp dev77.fastasa.no:mydata.dat ./

same thing — it is enough just to indicate the directory, and the copied file will be named the same as the original, by default

• devcmb> scp kasundev@dev77.fastasa.no:cvs/se/test.xml ./

gets the file from my “cvs/se” subdirectory (relative to kasundev’s home directory on dev77)

Putting an entire directory

• devcmb> scp -r my_source dev77.fastasa.no:Mysrc/bck/

copy directory “ my_source ” recursively (i.e., the directory and all its contents) from devcmb into a new directory “bck” in your home directory on dev77 (the directory “Mysrc” must already exist).

(Anyway I prefer to create a tar file and copy that files to the destination and unpack it … This is more clean.. I guess..  )

Getting an entire directory

• devcmb> scp -r kasun@dev70:MySrc/bck ./my_src/

copy directory “Mysrc/bck/” recursively (i.e., the directory and all its contents) from dev70 (where you are “kasun” into a new directory “my_src” in your current working directory on devcmb.