> Oh, dpb! > You remind me of Don Quixote tilting at windmills. > TMW is never going to change this. > We must adapt.
Indeed, I know it'll never change in the existing functions; one would hope that eventually the light would dawneth on the enchancement, though.
Just think what a pleasure it would be for to be able to throw away repmat('%f', ...) in favor of a repeat field as a side benefit! :)
> In practice, what I do is write a Fortran program that I run outside > Matlab to re-write the file in a format Matlab will accept. > If mex worked with gfortran, I'd do it within Matlab. > Alternatively, you can import the file into Excel with fixed format and write it out as .csv or .xls.
That's fine if one has the source or the flexibility to rewrite the Fortran output--if it is legacy code or has other restrictions; "not so much". But again, why in this day in age should it be necessary to adapt to the computer instead of the computer adapting to the user?(*)
And introducing another abomination of Excel simply as a transport interface for something as simple as a fixed-width formatted file???? Puh-leeeeze! :)
(*) Oh, I now see you're saying write a filter in Fortran -- yeah, that's a way albeit again it shouldn't be necessary. I've mentioned in the past I had written a mex-function that did at least a limited amount of interfacing allowing one to pass in a FORMAT string but I seem to have lost the source in the retirement move back to the farm and I don't have the ambition/drive any longer to try to reinvent it, unfortunately. You'd not think it would be that big a deal for a TMW employee...