Re: Question regarding user expectance.

["Florian Reuter" <freuter@novell.com>]

Hi Don,

thanks very much.

This really helps me evaluating the requirements for the new ODF1.2 numbering currently beeing discussed in the TC.

~Florian


>>> <donald_harbison@us.ibm.com> 03/23/07 2:30 PM >>>
Hi Florian,

I wish there was a simple answer to your question. As you know, MSFT is 
consistently defining interoperability as fidelity, where fidelity means 
the document presentation and content match regardless of the application 
translating, opening and saving a document file. Of course, they are only 
really focused on that problem within the universe of multiple 
implementations of MS-Office over the past 15+ years. If they really cared 
about delivering high fidelity translation between MS-Office and ODF 
implementations, they would be much more forthcoming with exposing the 
internals of their binary document file format technology, and they could 
even join the ODF TC and tackle the problem proactively. One is best not 
to hold one's breath on this account.

For customers challenged to implement open standards compliance policies; 
e.g. Massachusetts, Belgium, etc. we hear a more pragmatic requirement. 
For example, I spoke with the project leader responsible for the rollout 
of the ETRM (Enterprise Technical Reference Model) at Massachusetts state 
government yesterday. They are testing and piloting both the Sun and MSFT 
conversion utilities and reporting favorably with respect to the Sun 
solution. I asked him about expectations regarding fidelity. He said that 
they expect to see the content of the document preserved but they are more 
realistic regarding expectations of fidelity. 

One example he gave was the related to the standard letterhead template 
they have in MS-Word. When converted to OpenOffice the frames containing 
the graphic elements become compressed, and the appearance is distorted. 
They resolved the problem and it is therefore not an issue. 

Regarding pageination, numbered lists, customers expect them to 
correspond, particularly significant document-intensive professions; e.g. 
lawyers drawing up contracts, red-lining and distributing. 

From other feedback I've received, it seems that pageination variations 
are more tolerated. 

We did receive this feedback from Belgium recently:

"The Belgian directives say that by October 2007, ...  every federal 
public service will provide ?reading? functionality for the agreed upon 
format ....   This means that, if we want to have ODF being activated, the 
large majority of MSFT Windows Office users should be provided with code 
such that they are able to read ODF generated documents, spreadsheets, 
presentations.  "able to read" means here that the document one receives 
is identical to the document that has been sent.

Tests have been done on a set of 50 documents, but NONE of them has been 
satisfactory.  CleverAge, solutions from the ODF-alliance,  solutions from 
SUN,  in all tests that have been done, they come to major deviations 
between the original- and received document. " 

So you see Microsoft advocates have set the expectation for 'identical'. 
This indicates that they are taking a position that is much more rigid 
than the one I expressed above from Massachusetts.

This subject warrants some discussion and analysis in the Adoption TC, for 
sure. We may decide to collect as many views as we can find that are 
representative of a spectrum of expectations. Write them up, explain the 
difference between a standard specification, and its interpretation by 
application developers, and move to the topic of interoperability between 
these multiple ODF solutions *before* we add the more politically charged 
discussion of document interoperability between these ODF solutions and 
the dominant solution, MS-Office.

Best regards,

/don


Don Harbison
Program Director, IBM ODF Initiative
Business & Technical Strategy
IBM Software Group
tel:1-978-399-7018
Mobile: +1-978-761-0116
email: donald_harbison@us.ibm.com





"Florian Reuter" <freuter@novell.com> 
03/23/2007 08:54 AM

To
<odf-adoption-comment@lists.oasis-open.org>
cc
<zander@kde.org>, <office@lists.oasis-open.org>, 
<erwin.tenhumberg@sun.com>, <donald_harbison@us.ibm.com>
Subject
Question regarding user expectance.






Dear Adoption TC members,

I write to this list to get some feedback about question which arose in 
the OpenDocument TC.

Followed by a long discussion about text:numbered-paragraphs/text:lists 
the following question was raised by one of the
OpenDocument TC members Thomas Zander who represents KWord:
http://lists.oasis-open.org/archives/office/200703/msg00252.html

The interesting question is
<quote>
Basic question;
Given two different implementations of ODF, which have different ways to 
internally do numbering. Are they in violation
of the spec if they show different numbers for the list-items ?
</quote>

My initial response was that the answer is YES; the users want to have the 
same numbers in different applications.
However after thinking a little bit about it I found this problem very 
much related to the user requirement that e.g.
page breaks should be the same across applications --- which basically 
means that ODF must describe exactly how page
breaks are computed. (
http://lists.oasis-open.org/archives/office/200703/msg00253.html)

So is there some feedback from the Adoption TC; especially regarding user 
expectance? And maybe a little advice to the
OpenDocument TC?

Thanks very much,

~Florian