[Nexus-developers] NAPI state

[Peterson, Peter F.]

Actually, "NICK RULES!!!!" But that is beside the point.

I believe it is within Nick's authority to decide whether a particular
decision is made by him or should go by a vote. I don't have a strong
opinion on this, but if it is deemed necessary I'd recommend that the
tech committee creates it own rules for operating and posts them on the
NeXus website.

On the topic of hiding NAPI specific attributes. There are a couple of
unix principles
<http://www.catb.org/~esr/writings/taoup/html/ch01s06.html> that are in
direct conflict here, specifically "(7) the rule of transparency" and
"(10) the rule of least surprise". Least surprise would dictate that the
NAPI gives a consistent view of a file no mater what base is used. This
means that you should either have the type returned as an attribute for
HDF4 and HDF5 based files, or have it hidden in XML based files.
Remember that the NAPI is meant to give a consistent interface to all of
these file bases.

About the target attribute: I don't recall what this does, but vaguely
recall it is involved in links. Could you please explain what it is/does
so people (like me) can make an informed opinion?

P^2

-----Original Message-----
From: nexus-developers-bounces at anl.gov
[mailto:nexus-developers-bounces at anl.gov] On Behalf Of Mark Koennecke
Sent: Friday, May 20, 2005 10:00 AM
To: nexus-developers at anl.gov
Subject: [Nexus-developers] NAPI state

High,

the autoconf stuff has improved much, thanks to Freddy, and is good now.

Some tweaking with the text test
in the test case may be required.

Concerning the type: well it is extra work to rename type to NAPItype
and suppress it. But I see a reason for it to be done. 
On the  other hand, a guiding principle of unix is transparency and this
is why I am not so happy suppressing the type or NAPItype. I do not have
strong feelings about this and if we push back till june 3 it can be
done.

A bigger problem is renaming the target attribute: This touches all
API's and I wonder if we do need a vote in order to make such a change.

This brings up an organisational issue: Many decisions we make are
purely technical. and thus not interesting. But the things mentioned
above touch the content of the file. I personally feel that such
decisions should be made by a larger group. How to we come to a decision
anyway? Nick rules?


                                           Mark



_______________________________________________
NeXus-developers mailing list
NeXus-developers at anl.gov
http://www.neutron.anl.gov/mailman/listinfo/nexus-developers