TechLog

Ever wondered how to start a download from server at night from your laptop and shut it down with out disrupting the download?

Of course there are many ways, use RDP etc but screen is a better choice because its powerful.It gives the power of resuming from where we left.

http://www.howtoforge.com/linux_screen

Tuesday, April 19, 2011

C++ Singleton Pattern

C++ and the perils of Double-Checked Locking (DCLP)

http://www.aristeia.com/Papers/DDJ_Jul_Aug_2004_revised.pdf

Singleton: How should it be used

http://stackoverflow.com/questions/86582/singleton-how-should-it-be-used

C++ singleton design pattern

http://stackoverflow.com/questions/1008019/c-singleton-design-pattern

Friday, April 02, 2010

Perl HowTo: Parsing command line options

Sample program to work with CLI options. Play with this program options and pay
attention to the reported usage errors to better understand how GetOptions() work.

# CODE STARTS HERE

#!perl -w

=head1
 Using GetOptions() to process CLI options
 Ref: http://perldoc.perl.org/Getopt/Long.html
=cut

# Required module
use Getopt::Long;
use strict;

my $grade = 'A';
my $pcile = 0.0;
my $name = '';
my $result = 0;    # Fail by default
my $promote = 0;
my @marks = ();
my %history = ();

sub show_usage()
{
   print "\nUsage: --name name --grade [A|B|C|D] --marks m1 m2 m3\n" .
                "       [--pass] [--promote | --nopromote]\n\n";
   exit( 0 );
}

if ( !scalar( @ARGV ) )
{
   show_usage();
}

if ( !GetOptions( 'name=s'       => \$name,       # =  - mandatory, string
                  'grade:s'      => \$grade,      # :  - optional, string
                  'percentile:f' => \$pcile,      # read float value
                  'marks=i{3}'   => \@marks,      # sequence of 3 numerals
                  'pass'         => \$result,     # optional, flag
                  'promote!'     => \$promote,    # -promote and -nopromote
                  'history:s{2}' => \%history ) ) # sequence of 2 strings
{
   show_usage();
}


print "\n    SUMMARY   \n\n";
print "Name      : $name\n";
print "Marks     : ";
foreach( @marks )
{
   print $_ . " ";
}

print "\n";
print "Grade     : $grade\n";
print "Percentile: $pcile\n";
print "Result    : $result\n";
print "Promote?  : $promote\n";
print "\n";
print "History : ";
foreach( keys( %history ) )
{
   print $_ . " = " . $history{$_} . "; ";
}
print "\n";

0;

# CODE ENDS HERE

Wednesday, February 10, 2010

vim command to delete lines

My fingertips know the command to delete all the lines starting from current line till EOF (that's d + shift g) but I rarely remember the command to delete all the lines up to head of the file.

Luckily, vim offers a simple approach to delete desired lines, all we need to be aware is the following

. (dot) - represents current line
$ - represents EOF
syntax: start_line_num,end_line_num vim_cmd

Some example usages (Note: these commands are run in the editor, hence the beginning colon)

Delete all the lines starting from current line : .,$ del
Delete lines starting from 1 to so far : 1,. del
Delete all lines between 10 and 25 (inclusive) : 10,25 del
Delete all lines between 10 and 25 (inclusive) : 10,+14 del
Delete next 15 lines including the current : .,+14 del
Delete last 15 lines including the current : .,-14 del

Tuesday, November 03, 2009

Open Source Projects and Volunteers

Why only a handful of projects attract attention of developers/volunteers?

Moshe Bar and Karl Fogel in their book Source Development with CVS points that In a system that relies largely on volunteer energy, convenience is not a mere luxury—it is often the factor that determines whether people will contribute to your project or turn their attention to something with fewer obstacles to participation. Projects are competing for volunteer attention on their merits, and those merits include not only the quality of the software itself, but also potential developers’ ease of access to the source and the readiness of the maintainers to accept good contributions.

Point taken.

Saturday, October 10, 2009

Coding standards do no harm

Coding standards aid in picking up new code faster, it can help in debugging too. For example it's almost safe to skip a const qualified method while debugging a problem (just make sure that there are no mutable members).

Lately I had crazy time debugging a problem that eventually boiled down to non adherence of coding standards. Here it goes ...

A new API that looks as simple as below was added to a self sufficient component which compiles clean after the change was made.

    int process_data( const char * str, char delim = '  ' );

Suddenly another component that depends on the modified header fails to compile with error C2143: missing ')' before 'string'. That's a bit of shock!. By the way the other component doesn't use the new API at all.

It turned out that other component has defined a macro which is neither long nor uses UPPER_CASE. Of course it's their choice!

    #define delim TEXT(" ")        // and
     #define TEXT( str )  L##str    // just token pasting ...

You are punished for using a simple looking identifier (delim) and it takes efforts to understand that you are being punished!

Saturday, August 01, 2009

Defining INT_MIN

It's always better to take a hard look at warnings. Warnings are potential Errors. Recently I came across an interesting warning with VC++.

warning C4146: unary minus operator applied to unsigned type, result still unsigned

A bit of googling revealed that above warning can be dangerous. To understand the underlying problem, here is my own version of INT_MIN.


    #define INT32_MIN -2147483648

As 2147483648 is greater than max 32-bit signed int value (MAX_INT), it is treated as unsigned int. With this, type promotion kicks in and would fail the following condition.


    if ( 1 >  INT32_MIN )  // comparing two unsigned values
        std::cout << "I am sure, 1 gt INT32_MIN";
    else
        std::cout << "Surprse!  it says, 1 lt INT32_MIN";

To correctly define our own version of INT_MIN, see how it is defined in limits.h. The trick is to not to let the value to cross (signed) 32 bit int limits. The correct version should look like below.


    #define INT32_MIN (-2147483647 - 1)  // don't forget to put the braces around

Refrences:

1. More detailed explanation is available here
2. warning C4146

Tuesday, June 23, 2009

Quote from IBM GTO 2008

Good quote from IBM's Global Technology Outlook 2008.

A computer, like anything else, works best when it is built and used for a specific purpose. Though far more complex than a hammer or saw, a computer is a tool just the same. And all tools must be designed to a task.

Sunday, February 08, 2009

exstr dupms core on lenghty strings

Make messages normally uses exstr to extract strings that needs to be localized. It looks like exstr suffers from buffer overflow vulnerability. It dumps core on Solaris 9 with the following program snippet. exstr doesn't seem to process a lengthy string.


void PrintUsage()
{
    std::cout << "#mycmd -option1 -subopt1 -subopt2 -suboption3 \n\
                         -option2 -subopt1 -sbuopt2 -subopt3 \n\
                         ......................... \n\
                         ......................... \n\
                         -option15 -subopt1 -subopt2"
              << std::endl;
}

In order to generate message strings, I had to break the above snippet into ugly looking pieces - quite bad, I had to split the options mid way.


void PrintUsage()
{
    std::cout << "#mycmd -option1 -subopt1 -subopt2 -suboption3 \n\
                         -option2 -subopt1 -sbuopt2 -subopt3 \n\
                         ......................... \n\
                         .........................\n\
                         -option9 -subopt1 -subopt2";

    std::cout << "       -option10 -subopt1 -subopt2 -suboption3 \n\
                         -option11 -subopt1 -sbuopt2 -subopt3 \n\
                         ......................... \n\
                         .........................\n\
                         -option15 -subopt1 -subopt2";
              << std::endl;
}

Tuesday, January 06, 2009

VIXIMO's VixML platform for iphone

Looks cool, featuring

- Physics engine
- 2D and 3D visual effects
- Easy to use by non programmers

Tuesday, December 09, 2008

Zoetrope, new concept aimed at providing access to temporal web content. It lest people see how things (any) have evolved over a period of time. One useful thing that can be done using Zoetrope is to spot the best time to buy books on Amazon :-)

Zoetrope in action

Tuesday, November 18, 2008

Really cool stuff

Usability is still an issue with PCs. There are smart people (who works around an issue instead of simply nodding in disgust) out in the world ... proof? check this video

Tuesday, September 02, 2008

Google Chrome

Wish, it comes out soon :-)

http://blogoscoped.com/archive/2008-09-01-n47.html

http://blogoscoped.com/google-chrome/

Un named structures and VS8

With two un-named structures , the linker on Windows has thrown the following error

unnamed_struct.obj : fatal error LNK1179: invalid or corrupt file:
duplicate COMDAT '??1@@QAE@XZ'
NMAKE : fatal error U1077: '"C:\Program Files\Microsoft Visual
Studio 8\VC\BIN\link.EXE"' : return code '0x49b'
Stop.

The above is observed even with the varying number of struct members.

Known Issue: If two unnamed structs both declare a method with the same signature
and both are referenced, the compiler generates the same signature for both methods.
The linker then flags the .obj file as invalid due to duplicate COMDAT records
(More about this issue here)

The solution to this problem is, not to use more than one un-named structures