On Thu, 24 Aug 1995 18:52:15 -0400 Ken Paul said:
>Several people have responded to this thread and I think that it is
>time for the HP3000-L advocacy group to see if a change is needed.
>Almost a year ago, August 29, 1994, John Sullivan put out the first
>post regarding HPDesk not working with DDX.  Jeff Kell responded
>the next day to say that he thought it was probably the fact that
>HPDesk was using DBInfo mode 202 which still returns the current
>capacity and not the Maximum capacity.
 
And regrettably, I forgot all about this when Interex 95 rolled around
and even more regrettably while sitting in the database roundtable.
 
>It is now a year later and we are finding out that HPDesk will not
>support DDX (thanks for the info John) and that Jeff Kell and I are
>both calling for the change to DBInfo mode 202 to return the
>Maximum capacity instead of the current capacity so that HPDesk
>could take advantage of DDX.
 
>Let's now use the list to have a constructive debate on pros and
>cons for making this change and register your vote and then we'll
>see if we need to go to SIG IMAGE to ask for a change.
 
I can only see ONE supporting reason for not returning the maximum DDX
capacity for mode 202 -- if DDX is invoked and there is insufficient disc
space to accomodate the new extent, the dbput() will fail.  However, the
"enhanced" mode 205 probably doesn't check for adequate disc space either,
so a "truly" paranoid program would stick with the traditional mode 202.
In effect, mode 205 does no good except for a non-paraniod program in the
first place (which will NOT bother calling dbinfo() in the first place).
 
I would suggest that mode 202 returns the max DDX capacity, while mode 205
is altered to return the max capacity relative to available disc space.
Yes, that will be one whale of an overhead penalty, but it provides the
only reliable solution for a "truly" paranoid program.
 
>Maybe a year from now we wont have to have the same discussion.
 
Agreed, and shame on all of us for forgetting it at Interex.  I hate to be
guilty of lending legitimacy to the old "Ignore it and it will go away"
problem resolution philosophy.
 
>Ken 'trying to go through the proper chanels' Paul
>a.k.a. [log in to unmask]
 
Thanks for re-iterating this Ken.  Hopefully it will be etched on our minds
when IPROF rolls around (six months, not a year from now).
 
[\] Jeff Kell <[log in to unmask]>