= Preamble =

There is considerable interest in identifying FOSS equivalents of proprietary softwares. While there are many websites available with detailed information on such FOSS products, this page will focus on queries for FOSS equivalents received at NRCFOSS and the progress of the process set up to respond to such queries. The page will also provide details / links related to the migration related tasks undertaken.

== Coordinators ==
Dr. S. Srinivasan        NRCFOSS      srinivasan@au-kbc.org [[BR]]
-- one or two volunteers -- please put in your info

== Queries ==
1. Open source equivalents for Auto CAD & Photoshop

List of proprietary products & brief data  ; List of equivalent FOSS product & brief data [[BR]]
Relevant web site references

0. All rendering and postproduction work is done on Linux Cluster Servers as windows cannot handle this

1. Qcad is a 2D CAD package similar to Autocad.Its definitely a strong FOSS alternative to Autocad. [http://www.qcad.org]

2. There seems to be a lack of gui based mechanical 3D CAD packages in FOSS.Using a command line to create a 3D shape seems kind of strange.Also extremely difficult to learn.Brl-cad is one such package.[http://www.brlcad.org] 

3. For movies, animation and postproduction, the combination of the following packages has been successfully used:

Blender  http://www.blender.org/cms/Home.2.0.html

Gimp     http://www.gimp.org/

OpenEXR  http://www.openexr.com/

Verse    http://uni-verse.org/

Cinepaint  http://www.cinepaint.org/

Inkscape   http://www.inkscape.org/

These packages have been used to create an animation movie: http://orange.blender.org/

A good foss replacement for Matlab: http://www.gnu.org/software/octave/about.html

From a mailing list:

Octave is more robust but has a command-line interface and a sharper learning curve. It interfaces with gnuplot to produce graphics - itself not easy to use beyond simple ine graphs...:)

Scilab is faster to learn, easier to use, has a GUI interface, very neat demo programs, and can scale across compute-nodes using PVM. 

Both programs use numerical processing algorithms, though some engineers I know, feel that Scilab sometimes give inconsistent results.

We use Scilab for interactive stuff, and Octave for batch-processing, or as embeded within our protocols. (Actually - we use Perl wrappers for both R and Octave when we have to automate processing protocols for large data sets.)