Making Our Own AttributeProvider in OpenDtect

As the default AttributeEngine's Provider class cannot altered as we wish. We need to do some re-routing of it if our plugin wanted to access some protected or read-only members from the original Provider.

Here's the diagram showing what I alter to allow my plugin to do things that the default Provider won't allow it to do:



Something like that...

OpenDtect AttributeEngine Process Flow

In making attribute computation program using OpenDtect we should follow the right flow so that our algorithm will be run nicely.

Process steps:

1. prepareForComputeData First to be called, and only once
2. getInputData
3. computeData a const (called n times; depends on our processor core number ? and if we set allowParallelComputation to yield true)
   
4. finalizeCalculation

Again. This is a quite simple problem, but it takes some times for me to figure it out (d'oh)
NOTE: getInputData(), computeData(), and finalizeCalculation() are called n times, where n = number of traces in a line. (Called inside doWork() method of Provider class)

Spoiler Is Dead

After applying a new template.... my spoiler is dead...

Enjoy

Cleaning process for OpenDtect Objects

Previously: there is a pmake configuration to make OpenDtect Plugin...

If there's any change to the SRC.cc list then we should clean some objects created before, or those unused objects will make our plugin malfunctioned. (in other word: OpenDtect couldn't load the plugin)

Now, I just use this script (executed exactly on our plugin folder ? ):

#!/bin/bash
make source1.rmlib
make source2.rmlib

#and so on..

Simple... but without knowing this, making a plugin can give you (a pretty bad) headache!

oh c++...

Everyday Pmake Makefile for OpenDtect Plugin

SRC.cc := source1.cc source2.cc
PLUGIN := yes
PLUGINDEP := MySourcePlugin

OWNC++FLAGS := -DYOURFLAG1 -DYOURFLAG2

MODDEP := uiODMain
include make.od.Defaults
include make.Targets

Ubuntu??

Ubuntu is Linux

OpenDtect (Preparing & Deploying Plugin)

Preparing:

1. Always use function below (XXX is your plugin name):

#include "plugins.h"

extern "C" int GetXXXPluginType()
{
    return PI_AUTO_INIT_EARLY;
}

extern "C" int GetXXXPluginInfo()
{
    static PluginInfo retpi ={
        "Plugin Name",
        "Author",
        "Version Number",
        "Description" };
    return &retpi;
}

And don't forget the init function!!!!

(Part Of) Deploying:
2. Setting *.alo files
Place your *.alo files on :

• The plugins/platform_dir of the installation directory ($DTECT_APPL/plugins/$HDIR)


• The '.od' directory in your 'Personal directory' ($HOME/.od/plugins/$HDIR on Unix)

Example of: odmain.xxx.alo <-- this is the "default" file naming format

Annotations
Madagascar
uiMadagascar
CmdDriver
GMT
uiGMT

3. Distribute your plugin files (those *.dll or *.so files)

Further Resources:

• Opendtect Plugin Guide


• The Source Code

OpenDtect (Linux Development Install)

Hi...

1. Install GCC v3.4.6 (g++, gcc-base, cpp, gcc, libstdc++6-dev. ubuntu, hardy)


2. download opendtect binary package and developer source package here


3. chmod downloaded files to executables, run the file, it will extract *tar.gz file into the same directory

   chmod +x [filename]



4. extract to home directory (e.g: /home/someone)

   tar -xf [extracted downloaded files]



5. Install opendtect, setup development environment from

   Utilities > Tools  > Create Devel. Env.



6. Setup IDE (in my case eclipse CDT), Create a new c++ project with existing source, pointing to ODWork directory.


7. Start developing! in plugin folder of ODWork

Css Class for Prettyprint

class="prettyprint lang-sh"

Source: google "prettyprint"

Old GCC (Ubuntu Releases) & How to add them (on Later Version of Ubuntu)

To Install: (currently trying Hardy's release)
gcc-3.x.x
- gcc-3.4.x-base
- cpp-3.4.x
- gcc-3.4.x
- g++-3.4.x
- libstdc++6-dev (=3.4)
Note

1. Hardy Devel <- Use This


2. Hardy-update Devel

Adding repository & Installing (GUI way, using hardy package):
1. software-properties-gtk &
2. add deb http://ubuntu.pesat.net.id/archive/ hardy main universe
3. Update
4. synaptic &
5. Search & install g++-3.4
6. Done

Some extras: (earlier package search)

1. Ye Olde Ubuntu Releases (Headers, No Hardy)


2. Not So Olde Ubuntu Releases (ISO)

Parallel C#

Parallelcsharp.com

INSTALLATION (Click To Show)

2. INSTALLATION
================

1. Untar the archive into the folder `~/pcsharp' (or any other folder)

2. Compilation (It`s supposed that pcs (Mono) and gcc (C++ compiler)
had already been installed on your system).
cd ~/pcsharp
make

3. Add the following environment variables:

$PCSHARPPATH - the main directory of Parallel C# is located here
$MONO_PATH   - Mono will search for assemblies in these directories
$PATH        - add here the directory `~/pcsharp/bin'

If you use bash, you can do it by adding the following lines 
into files `~/.bashrc' and `~/.bashrc_profile' in the following order:

export PCSHARPPATH=/home/yourhomedir/pcsharp
export MONO_PATH=$PCSHARPPATH/bin:$MONO_PATH
export PATH=$PCSHARPPATH/bin:$PATH


3. Runtime Initialization
=========================

This version of runtime uses `rsh', `ssh' or possibly other types of rsh-compatible
remote shells for initialization on the nodes of cluster.
rsh/ssh must be configured for transparrent logging into the nodes of the cluster
from frontend.
There`s a `pcsboot' utility in this package for cluster initialization.
It has only one optional parameter - the name of the file, that contains
the names of the cluster nodes (each node must be on the separate line).
File `$PCSHARPPATH/bin/nodes' will be loaded when this parameter is omitted.

Usage example:
 pcsboot $PCSHARPPATH/bin/nodes
or
 pcsboot

4. Finalization of Runtime
==========================

When you finish your work with the cluster, you should finalize the
Runtime on the nodes. Command `pcshalt' is dedicated for this task.

5. Additional Tools
===================

To find out which nodes are being used by Parallel C# Runtime currently use `pcsnodes' command.
It prints the list of nodes.

To execute some task on all nodes of your cluster use command `do-all'.
For example:
 do-all killall -9 mono

This command will kill all mono applications on your cluster.

Note: `do-all' utility executes the same command on all nodes that are listed in file
`$PCSHARPPATH/bin/nodes'.

6. Compiling your Parallel C# programs
======================================

In this package there`s a compiler called `pcsc' for
compilation of Parallel C# programs. There`re two stages of compilation:
- Parallel C#-code is translated into C# code
- C# code is compiled by `gmcs' compiler (from Mono platform).

Note:
 all parameters for `pcsc' compiler are passed to C# compiler (gmcs)

Examples of usage
 pcsc fib.pcs
 pcsc fib.pcs /out:myfib.exe
 pcsc /t:library fib.pcs

You can compile Parallel C# programs into DLLs
 pcsc a.pcs /t:library /out:a.dll

Then you can include it into the program
 pcsc /r:a.dll b.pcs

You can mix Parallel C# and C# files
 pcsc a.pcs b.cs

Or compile several Parallel C# files at a time
 pcsc a.pcs b.pcs

If you need only to translate Parallel C# files into C# code you can use option `/parseonly'.

7. Starting your Parallel C# programs
=====================================

Local mode:
 mono fib.exe 35

This will start program in local mode, i.e. movable methods will be executed as new threads.
No network operations will occur.

Distributed mode:
 mono fib.exe 35 /np 35

This will start program in distributed mode on 35 processors.
In this case movable methods will be transferred across the network.

Showing NetworkManager Applet in LXDE (Ubuntu 10.4)

Langkahnya:
1. Alt+F2
2. Masukkan perintah nm-applet
3. Enjoy