Re: Loading of Matlab 5 Files

[Andy Adler]

>| MH> MATLAB v5 can save data in a v4-compatible way, if you don't use
>| MH> the new features of v5.   
>| MH> From help save: 
>| MH> SAVE fname X Y Z  -v4  saves a MAT-file that MATLAB 4 can LOAD.
>| 
>| Thank you. I know this option. The reason of my question was that the
>| exchange of data with Matlab users becomes more difficult, when you
>| have to ask them to save the data in a special way. So, if octave
>| still wants to be matlab-compatible in the way that I can write programs
>| which can be used in octave as well as in matlab, then the
>| implementation of the new binary format seems to be necessary.
>
>Are the specs for the Matlab 5 binary file format publicly available?
>If so, would anyone be interested in writing the code for Octave?

It doesn't appear to be that way...
I just looked at http://www.mathworks.com/support/solutions/v4/7279.html
_____________________________________________________________________
Problem 
Is there any documentation on the structure of  MAT-Files in MATLAB 5?

Solution 
Currently, there is no documentation available regarding the MATLAB 5
MAT-File structure. There are no plans to publish this information.
We recommend using our MAT routines for creating MAT-files in MATLAB 5.

Please note that MATLAB 5 does still read MATLAB 4.x MAT-files.
Therefore, if you have code that you used in MATLAB 4.x to create
MAT-files, you can still create MAT-files with this code and read the
MAT-files into MATLAB 5.

You will not have access to the new data types supported in MATLAB 5
if you use your old code.  For this, we recommend using the MAT
routines provided in the MATLAB Programming Interface.
[ cut ]
Date Last Modified: Jul 31 1997 
_____________________________________________________________________

<rant>
This really offends me. I'm not as diehard a free software fan
as some. But, after all, it is _MY_ data. If I save data with Matlab,
then they got me locked into buying Matlab to access my data.

Sorry for whining in the forum. I'll go and whine on
comp.soft-sys.matlab as penance.
</rant>

I would offer to take on this project. But unless Matlab changes their
mind... there's not much I could do. On the other hand, we could
try to reverse engineer the interface, but then we're not garuanteed
that what we have won't change with the next version.
_____________________________________________________________________
Andy Adler,    | Pulmonary Physiology Unit         | Lab 303-398-1626
address@hidden | National Jewish Center,Denver,USA | Fax 303-398-1607

   For the Snark _was_ a Boojum, you see.    -Lewis Carroll