[HEAD branch] a couple of advices

[Luciano]

Hello,=20
I'm a bit further in my quest for a compilable HEAD.
However, the changes that I'm about to commit are a bit less trivial than=
=20
those I submitted earlier, so I would appreciate some suggestions.
Here are the changes I'm about to commit:

* the new khtml code needs the interfaces/kio/html.h include file.
  The include file is pretty trivial, containing only an enum.=20
  My change involves adding the interfaces directory to CVS
  under kdesrc, and a Makefile.am with the single line=20
  SUBDIRS =3D kio=20
  in it, and a change in the prepare_kde_src script to copy the needed fi=
les
  in the kdesrc/interfaces/kio subdirectory. Is this the correct approach=
?
  I don't like adding more stuff than necessary, but this seems necessary=
 to   =20
  me.
* the second change is needed to compile th kssl subdir.
  As it is now, the original configure.in.in is copied by the prepare_kde=
_src
  script, but in it the AC_OUTPUT(kio/kssl/ksslconfig.h) is commented out=
 and
  furthermore it creates the file in the wrong place. So I'll add=20
  configure.in.in to CVS and remove the 'cp' from the script. This is=20
  annoying, since this way changes to the original config.in.in should be
  tracked, but on the other hand, we are already cherry-picking the sourc=
e  =20
  files of the kssl directory.

An issue I really need advice on is about the kfileitem.h in the dropin/k=
io
directory. It redefines UDSEntry, which is also defined in kdesrc/kio/glo=
bal.h
The problem is that this time kfileitem.h and global.h are both included =
by=20
some files, so a comflict happens. As a temporary workaround, I have=20
commented out the kfileitem.h definition, and made the code related to it=
 a=20
NOP, however, I need to know what uses the kfileitem class. Worst case, i=
t=20
needs to be reimplemented with the global.h definition...=20

Thanks in advance for your attention.

Luciano Montanaro