RE: Which group PTFs do we need? -- MIDRANGE-L

Mark, I feel your pain, sort of.  I tend to agree, it's gotten much more
complicated to stay current with necessary fixes.  My understanding of the
concept of group PTFs was that they would be delivered quicker than they
could otherwise if they were including their contents in the cumes.

But you know what?  If I'm doing this, I'm not going to bother with group
PTFs until I install a cume anyway!  Look at the PSPs for the groups and
you'll see that you have to be at a certain cume level before you can
install the group.  If I were responsible for managing PTFs on our system,
in addition to all of the other tasks that go with that kind of a job, the
last thing I'd want to have to do is to go through the rigamarole you
described.  How do I know which ones?  Load each CD-ROM, one group at a
time.  Is sequence important, especially with the cume?  yada, yada, yada.

If you have issues between cumes, install specific individual PTFs and their
pre- & co- reqs.  Just like the good old days.

db

> -----Original Message-----
> From: midrange-l-bounces@xxxxxxxxxxxx / Mark S. Waterbury
> Sent: Tuesday, October 26, 2004 10:37 AM
>
> Hello, all:
>
> This whole discussion of "Group PTFs" begs the question, why do we need
> "Group PTFs" at all???
>
> I thought the whole idea for CUMulative PTF packages was to make
> it easy for
> customers to stay current, packaging ALL of the PTFs they need in one
> convenient bundle that could be installed as if it is one large PTF.
>
> But, over the past few years, it seems that IBM has corrupted that concept
> by introducing various Group PTFs (for WebSphere, for DB2, for
> Java, etc.),
> and this really complicates our lives -- trying to figure out
> which ones to
> order, and what sequence to install them, etc.  (And just try to
> explain all
> of this to a NEW iSeries customer!)
>
> Also, if I have the latest CUMe PTF package installed, do I really still
> need any Group PTFs? If so, why? I thought that was the whole
> point of CUMe
> packages?
>
> I was told that IBM supposedly did this (introduced Group PTFs) to reduce
> their costs and make it easier for customers to order only those PTFs they
> "need" (e.g. all PTFs for DB2 or all PTFs for Java), but in reality, since
> customers do not know which Group PTFs are really needed, they ended up
> ordering ALL of them (or all the ones they think they needed), so
> I suspect
> that IBM's actual costs to create CDs and ship them to customers were
> actually higher than staying with the original CUMulative PTF package
> concept.
>
> Also, now, with iPTFs and the ability to download the whole PTF
> package via
> the Internet, there really is NO cost for burning CDs or shipping, so that
> argument goes away also.
>
> Has anyone figured out a good strategy, other than to order ALL
> of the group
> PTFs that seem to apply?
>
> And, when you have to install more than one Group PTF, which one do you
> install first? The CUMe PTF package, then the group PTFs, or the other way
> around?
>
> If I order and apply ALL of the group PTFs, do I still need to order and
> install the latest CUMulative PTF package?  And, perhaps more
> important, if
> I install the latest CUMe PTF package, do I really need to order
> and install
> ANY Group PTFs? If so, why? I mean, isn't the CUMulative PTF package
> supposed to contain all of the latest PTFs that I need?
>
> I think this is one area where IBM has really gone "backwards"
> over the past
> few years, under the false assumption that this was "progress" --
> delivering
> PTFs to customers faster, etc. -- in reality, they have apparently created
> more problems for customers than they solved, IMHO.
>
> I think we were all far better off, and would be so again, if IBM
> would just
> get rid of the group PTFs and keep the CUMulative PTF packages more
> up-to-date, perhaps refreshing them monthly, since there is no longer any
> cost for burning all of those CDs, and shipping, etc. -- just let us
> download them, via the (relatively nice) iPTF process.
>
> That's just my opinion -- what do you think?
>
> Regards,
>
> Mark S. Waterbury