RE: [oic] odf 1.2 part 3, packages, features

["Dennis E. Hamilton" <dennis.hamilton@acm.org>]

Svante,

No I haven't tracked down any external sources of tests for technologies ODF
(Packages) depends on by reference.

The first thing I notice is that we need to profile what our normative use
is of particular external specifications and any additional conditions that
are placed on them for ODF (Packages).  Then we need to find out where there
are agreed tests that could be adapted to the permitted cases.  

I think that the SpecAnalysis Wiki pages can be used for collecting that
kind of information and tying to further materials in the OIC SVN and
elsewhere.  Fortunately, the ODF 1.1 Chapter 17 on packages is the
most-developed portion of the SpecAnalysis.

I am reluctant to put too much work into developing an ODF 1.2 SpecAnalysis
yet, because I am not sure how stable the organization is, even for ODF 1.2
Part 3.  There is probably some provisional structure we could create to
anticipate 1.2 without becoming too vulnerable to changes from CD to CD.

The features identification might be helpful, if we didn't lock it to the
organization but to the concepts.  It would also help identify and confirm
all material differences between the ODF 1.1 Package and the ODF 1.2 Package
specification.  It is valuable to know what the differences are, especially
any that might be seen as breaking changes or limiting backward
compatibility.

 - Dennis

<SOAPBOX (no need to comment on this, I just wanted to think out loud a
little)>

I don't believe that every flavor of Zip option is meant to be useable, for
example, but we need to at least comprehend the profile for OIC.  Likewise,
I don't believe every Zip compression method is meant to be used.  The
DEFLATE compression is documented pretty well and I believe there are
independent tests for it, just as there are "test vectors" to work against
various hashing and encryption methods.  Finally, there are issues around
how some of the external sources are themselves underspecified and how that
is resolved in practice.  PKWare AppNote 6.2.0 is a great example of that.

In some cases, there seems to be reliance on common libraries of operating
systems or development tools.  I have no idea how their conformance to what
they support is established or if it is even documented at all.  

Somehow, we need to reverse-engineer our way out of this, but it makes my
head hurt to think about it too often.

Fortunately, anything we do at OIC to improve the level of understanding
will be valuable in verifying things about ODF 1.1 implementations and
providing guidance, perhaps, to ODF 1.2 implementations if not the 1.2
specifications themselves.  I think there is time for some feedback to the
ODF 1.2 specs.
</SOAPBOX>


-----Original Message-----
From: Svante.Schubert@Sun.COM [mailto:Svante.Schubert@Sun.COM] 
http://lists.oasis-open.org/archives/oic/200912/msg00063.html
Sent: Thursday, December 17, 2009 09:31
To: dennis.hamilton@acm.org
Cc: oic@lists.oasis-open.org
Subject: Re: [oic] odf 1.2 part 3, packages, features

Hi Dennis

have you (or someone else) already investigated, if we could rely on exiting
tests from external technologies?

Regarding the package features, there are half a dozen
 
[ ... ]