ISOmorphin

It’s been a long time since last ticket on this blog.

But no, ISOmorphin isn’t dead. Development had been paused because I implemented all the functions that I needed.
So, what ISOmorphin lacks? For new functions, it belongs to you to say it to me. Otherwise, it needs testing to be more stable.

The last part of ISOmorphin which needs IMO development is conversion part. My software used iat as a GUI to do those things. The bad news is that in my distribution (Ubuntu), iat is packaged in a such way that I can control it. You will say, ask the packager to modify his work for suiting your needs. But iat is still in development, and a rewrite is beeing made by his author Salvatore Santagati. So I prefer waiting for iat 0.2 which will include faster conversion and VCD support before rewrite my conversion part.

So next version of ISOmorphin will include iat 0.2 control to convert images from notISO images to ISOimages.

Keep Tuned…

For now, “Media image to ISO” part is done by iat software by Salvatore Santagati. The problem is that the packaged version, even if it is called 0.1.3, is not the same that the version avalaible on the project site, at least for Ubuntu.
This last version is necessary to take advantage of a progress bar during the conversion of an image.

ISOmorphin uses at the moment a local version of iat which need to be compiled by oneself to use the conversion of CD images.

I thus decided to port the code C ++ of iat in Python to integrate it directly into ISOmorphin.

Once this have been made, I made a small benchmark to compare both versions, the C++ one of Salvatore and mine in Python. I so measured the execution time of the program to convert various files of very variable sizes. Every file was converted 5 times with every version of the program.

Tests were realized under Ubuntu 8.10 64 bits with a processor Athlon 64 3200 +, 1Go RAM and a generic Linux 2.6.27 kernel.
Least that we can say is that the results are unexpected:

Performances benchmarks Python vs C++

We can see there that the conversion is much longer in Python for small files, which is something what we could expect. But with bigger files (300 Mb and 4Gb), the conversion is faster in Python… It is necessary to note that, for the 4Gb image, the distances-types recover, thus it is impossible to assert that the Python version is faster.

Nevertheless my version does not seem to suffer from degradation of performances thus it is going to be includes in the next version of ISOmorphin, what will allow an easier distribution of the program.

Source code of both versions :

iat,  Python version

iat, C++ version

The site is on-line for more than 2 months now but I was not able to update it due to the lack of time.

Nevertheless the project advanced little because Seb0uil joined me. He filled the subversion repository on gna.

I also come to upload the version 0.2.1 in the download space.

Objectives for the next version are:

• Addition of a script of installation
• Addition of the support for backup of Audio CD

Various sites are speaking about ISOmorphin:

1. Ubuntu-fr forum where the project was launched,
2. Freshmeat on which I referenced ISOmorphin,
3. Ubuntu, carnet d’un débutant,
4. Packman where an OpenSuse package is available,
5. K-Hole, in italian,
6. Crismon’s Blog, in italian too,
7. Linux RSP, in russian,
8. Softpedia who added ISOmorphin yesterday in their base

To finish, a small call:
ISOmorphin needs testers and ideas to improve it. Indeed, the software works perfectly at home thus if nobody reports bugs or asks for new features, it is likely that the project will not move a lot…