[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]



I'm not talking about simply moving existing procedures...I'm talking
about a new SESSION module to encapsulate the MODS session

I went back and looked at FTPAPI, since it had been awhile...

It might be a significant rewrite to decouple the FTP procedures from
the MODS session implementation...though I have to think it would be

A shortcut might be to have a SESSION_GetInstance() at the beginning
of the procedures  to retrieve  the single instance a procedure needs
to work with.  Processes that update session variables could then call
a SESSION_SaveInstance() at the end to store the changes.

If I get a chance, I'll take a longer and harder look at what just
what would be involved.


On Mon, Oct 18, 2010 at 8:31 AM, Dennis Lovelady <iseries@xxxxxxxxxxxx> wrote:
> Thanks, Charles.
> That would seem cleaner - but the fact is, that’s what we have now, and as
> Scott indicted in the comments, the module is larger than we might lock.
> Maybe I'm missing something but I cannot envision a way to put all the
> references to MODS subfields into one module without limiting ourselves to
> only that one module, since the multi-occurrence variables are used all over
> the place.  (Consider, for example, wkErrNum, which is a subfield of MODS
> wkSession...)  So I'm interested - how would you pull this off?  (It's a
> serious question; maybe I'm missing something important.)
> The only cleaner-looking way _I_ can come up with for this, is to use
> pointers exclusively, especially if we're going to keep V4 compatibility.
> But that would move the initialization of the variables away from their
> definitions.  Maybe that's not such a big deal, but I think there's value in
> having the definition tied to the initial values.  And I really like that
> one RESET can set a slew of variables to their initialization values.
> Apparently someone else liked that, too.
> Dennis Lovelady
> http://www.linkedin.com/in/dennislovelady
This is the FTPAPI mailing list.  To unsubscribe, please go to: