[odf-discuss] Microsoft and Novell cooperate on document formats

[Ian Lynch]

On Fri, 2006-11-03 at 12:08 +0100, Thomas Zander wrote:
> On Friday 03 November 2006 11:37, Ian Lynch wrote:
> > > I doubt that, what do you base this on?
> > > Based on my years of actually programming on a WP app and seeing
> > > various items in the spec, 10 years sounds plausible.
> >
> > Impression Publisher on RISC OS I believe took a team of about 12, 2
> > years to development and that was written entirely in assembler
> 
> I gather you don't actually have significant experience in the field :)

Only as a user of many WPs and considering the most effective one I have
used was Impression Publisher even to-day. Although I have written
software in the past it was not in the field of office applications but
I don't consider that too relevant to this particular issue. It seems to
me that if OOo has developed ODF in less than 10 years it should be
possible to develop a single application that implements a subset of
similar XML based components in rather less time. If indeed Moox is so
much more difficult to implement than ODF, MOOX is a dead duck before it
even starts. I don't need to be a coder to work that one out.

> The spec we are talking about is many thousands of pages long with all 
> the bugs that were made in Office being put into the spec as well.

Well that is also true for reverse engineering filters for OOo without
the documentation as well as developing ODF and a whole host of other
things in a lot less than 10 years so I can't see why it would take 10
years even with these other constraints.

> Basically, there is nothing in the world (except for Office) that comes 
> close to being in the size and scope of a suite that would do this.

I did say it depended on what was meant by an app. Implementing the full
spec in all its parts in a cross platform office suite is a lot
different from implementing part of it in a simple standalone
application.

> It reads like you make a very human mistake of underestimating the work 
> you don't fully grasp.

Not really, its more that the problem is ill-defined. 10 years to
implement the spec in any app is open to wide interpretation. 10 man
years or 1000 man years? Without saying how many people are involved the
unit of measure is rather meaningless. One app and a subset of the spec
or only apps where the full spec is relevant? I think we are just
interpreting the problem differently.

So if you are saying fully implementing MOOX as a native format in OOo
or Koffice and all their components could take 10 years or never I'd be
ok with that. (could be the reason why MS don't implement odf
natively ;-) ) If you said implement a new basic text editing
application that uses a relevant subset of MOOX as its native format
given unlimited resources I'd bet my house on it that I could get that
done in less than 10 years.

Ian
-- 
www.theINGOTS.org
www.schoolforge.org.uk
www.opendocumentfellowship.org