RE: Source file names (was Using prototypes...)

["Dennis Lovelady" <iseries@xxxxxxxxxxxx>]

> On 11/05/2010, at 11:25 PM, Lim Hock-Chai wrote:
>
> > I wonder if it would be easier for IBM to simply increase the
> > 10-character limit to say 30-char.
>
> Pretty much too late to do that. The costs would outweigh any benefit.
> Rochester designed 30 byte names into the MI but chose to expose only
> 10.

I did not know that, Simon!  Interesting!

But this is a problem that MacroSloth solved umpteen years ago to get past
their 8.3 shortsightedness.  It was done with an additional layer of
indirection (yes, as always this at the expense of performance), and IBM
could choose to do the same thing.  I doubt it would be considered
worthwhile, but the infrastructure is already in place with the IFS if you
think hard about it.  Today we expose native programs (meaning traditional,
RPG et cetera) to IFS via soft link, such as Scott Klement's SQLQSH program,
et cetera.  There are also Java extensions in the tool box that extend Data
Areas and the like in a similar manner.  It doesn't stretch the imagination
too far to think that similar things could be done in that or a new file
system.

But the question: would it be worth it?  Not to me, certainly, as I've
become accustomed to the nuances and I do feel that the IBM standard of
naming suits most purposes.

Dennis Lovelady
http://www.linkedin.com/in/dennislovelady
--
"I get my exercise acting as a pallbearer to my friends who exercise."
        -- Chauncey Depew