Paul Blackburn wrote:
>Dear AFS Folks,
>
>It is important to talk with the _right_ people to get things done.
>
>Gary Gerchak is in DCE marketing. He is not in IBM Transarc Labs.
>I am not sure that the DCE marketing people would like to see AFS
>continue to grow and be such a success with customers.
Without drawing more names into this, Gary's name was given to me by one of
the front-line support people as the individual who made the decision in this
case. That does seem odd that someone in DCE marketing would be making
decisions regarding afs contrib. Also, in my conversation with Gary, he
placed himself squarely in the role of afs-policy-decision-making-authority.
Look, you're part of the evil empire. You tell me why I was directed (by the
'normal support channels') to someone in DCE marketing regarding a decision to
refuse something to afs contrib. (eh?)
<flagrant cheerleading deleted>
>IBM would be "killing the golden goose" if it did not listen to its AFS custom
>ers
>and respond to their business needs.
Well, in this case that's exactly what they are doing. (Also, see Atro
Tossavainen's post. I believe that his desire to see somthing contribbed that
meets his needs falls under the same situation as mine.)
>PS I know that the current release of AFS 3.6 for Linux runs with kernel 2.2.1
>6
>However, we will soon see a new release of RedHat and I hope Transarc will
>be able to update their AFS linux release for the newer kernels.
The problem here is that the supported product will _always_ be _months_
behind kernel development. People who choose to forgo support to get afs on a
particular kernel should have that option if someone else can make that module
available to them. Personally, I would like to see Transarc provide the
un-thoroughly-tested bits on an unsupported basis, with appropriate "This is
an upsupported product" spewed to your screen every time you do an open() call
or something. But given decisions like I found, that's a pipe dream.
Dave
