Re: Really? There's still people using the QDLS file system on IBM i 7.3?

On Fri, Jan 6, 2017 at 7:58 AM, Rob Berendt <rob@xxxxxxxxx> wrote:

Again, we're talking about people running into issues when they open a
share on IBM i and click on File, Copy and that aborts on the paste when
it tries to call it
copy of oldfile.doc
because that violates the 8.3 naming. I think that hardly qualifies as
"code that works".

But that's just it - in the scenario Scott is talking about, that
doesn't happen. What you're talking about isn't even "code" at all.
It's UI. Back-end *code* doesn't involve people "clicking" on things.
Or even typing copy commands at a command line for that matter.

I supposed one could keep redefining "code that works"
to keep clinging on to the status quo. It works "if" you follow this ever
tightening set of restrictions.

It's not ever-tightening. It's static. Scott is talking about (for
example) folks who have a fully automated process that happens to use
the /qdls folder instead of some other IFS directory. The naming issue
is baked into the code; the code knows to manage files as necessary so
that there aren't "copy of" situations or anything else that wouldn't
fly in QDLS.

Nobody is saying people should write new stuff that uses QDLS. But
there is such a thing as code that uses QDLS which meets the current
need, while other needs are more pressing. That's all Scott is saying.
He isn't even saying that people SHOULD ignore the QDLS issue in the
face of other needs, just that depending on your shop, it might be a
tough sell to address it.

John Y.