Re: V5R2: Source now allowed in the IFS

Scott

V5R2 has brought a new implementation for IFS. The old one had more layers
between the user and the data, so was slower. I think (might be mistaken
here) that is was based on dataspace technology, just like physical files.
The new one is faster and more space-efficient.

So, with the new storage type and variable-length records (CR-LF
delimiters, e.g.), there is great potential for saving space. But you have
to be on V5R2.

At 02:01 PM 10/15/02 -0500, you wrote:

Why do you think that the IFS is "SO inefficient a user of space"?

It seems to me that the fixed-length source member is inefficient, since
all the blank characters at the end of every line of text are wasted in a
fixed-length system.

So, I'm baffled by your statement.  Is there something about the IFS that
I don't know?  Something that makes it waste even more space than that?


On Tue, 15 Oct 2002, Al Barsa wrote:
>
> I was at the Inge Weiss session yesterday where she indicated that source
> can be put into the IFS, but the IFS is SO inefficient a user of space, why
> would I ever want to ever want to waste that disk space?  (PS:  This
> observation was taken in the V4R4 time frame/)
>