--===============1568758390==
Content-Type: multipart/signed;
	boundary="nextPart1444263.ffaApd1dgN";
	 protocol="application/pgp-signature";
	 micalg=pgp-sha1


--nextPart1444263.ffaApd1dgN
Content-Type: text/plain;
  charset="utf-8"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On Friday 27. June 2008 18:09:45 Stefan Nikolaus wrote:
> On Friday 27 June 2008 16:51:49 Thomas Zander wrote:
> > On Friday 27. June 2008 11:48:27 Stefan Nikolaus wrote:
> > > Hmm, you speak of making each table page derive from KoTextAnchor? Th=
is
> > > would result in a set of anchors and the user becomes able to insert
> > > text between the table pages. This should be avoided. Or did I missed
> > > an important fact?
> >
> > You need one instance of the anchor and one instance of a table shape f=
or
> > each page, yes.  Otherwise it won't work (things like clicking and
> > repainting of dirty areas need a shape).
>
> ... and the anchor to get the information about the text shape we are in,=
 I
> see. The text shape is a KoShapeContainer, but it isn't, because of
> representing a set of pages, is it? 'Just' because you can insert other
> shapes and anchor them? The page (or frame) creation/deletion is managed =
by
> KWord? I want to use the table shape as another view mode in KSpread late=
r.
> Just to mention it; that does not prevent using kotext, but relying on
> KWord's internals. :|

The text shape is a KoShapeContainer for no other reason then to be able to=
=20
have child shapes.  Like the table shape, or any other shape.
Without that feature the text shape could have extended a plain KoShape and=
=20
still be multi page etc.

KSpread can use the table shape later on and kspread would be responsible f=
or=20
setting the size and position of each shape (where I assume one sheet is on=
e=20
shape). The table shape should have a way to have the data model be shared=
=20
between different shapes.   All of this can be done without the table shape=
=20
being a shapeContainer.

> > I'm afraid that indeed the user could insert a character in between, I'm
> > not sure how to fix that.  This is indeed one of the reasons that I want
> > to have tables that are part of the text instead of drawn on top of the
> > text. i.e. use QTextTable.
> >
> > But I don't think its a showstopper that a user can insert something in
> > between, do you?
>
> Well, it's a bit odd. Even more, if we support breaking the table inside a
> row, i.e. the row's cell contents are painted until the page edge is
> reached and continued on the other page. But also with breaking at cell
> edges, it can be quite confusing. And what happens on laying out the table
> pages on position changes? I guess, any inserted text is moved behind the
> table.

I certainly would call it an unsupported 'feature' ;)

=2D-=20
Thomas Zander

--nextPart1444263.ffaApd1dgN
Content-Type: application/pgp-signature; name=signature.asc 
Content-Description: This is a digitally signed message part.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQBIZnw/CojCW6H2z/QRAk8bAJoDVh3eZ5UWl63D50/Ths+h2+wQyACgqMF6
aceCi7F6XKtRigFmh7Hd9kc=
=IqKV
-----END PGP SIGNATURE-----

--nextPart1444263.ffaApd1dgN--

--===============1568758390==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
koffice-devel mailing list
koffice-devel@kde.org
https://mail.kde.org/mailman/listinfo/koffice-devel

--===============1568758390==--