FW: Fwd: [dot-notify] When will Konqueror become the number two browser?

Ralf Habacker Ralf.Habacker at freenet.de
Mon Jun 30 22:27:59 CEST 2003

> Hi,
> so i will give you an update of what i am doing, and what work would have to
> be done by other people, which are more involved in this than me ;-).
> i am paid half time for working on open source stuff by automatix, the same
> way as zack rusin. so at the moment money is not the problem.
> the vision:
> there is no real reason, why kde should not run on ms windows, the problem is
> only, that nobody is working on it. so i looked into the sources somewhen in
> last october, and started fiddling around with this idea. and i didn't find
> any reasons, why this should be impossible.
> what i did/what already works:
> i started to port qt for x11 to windows gdi native, so that we can eliminate
> the X server, which is quite nice for "normal" X11 programs, but which only
> eats cpu and "catches" all the kde windows in itself, when working with kde
> programs.
> i tried to eliminate the X dependencies from kde 3.1, i worked on linux for
> that, and that works so far, that i can use some kde programs on the
> framebuffer. that is still work in progress, it is ugly and so on, i only did
> it to identify, which problem areas would arise by removing X.
> the following areas need to be fixed for that (not only commented out
> as i did
> to get it running):
> - the whole NET:: stuff for raising and lowering windows and so on
> - the keyboard input stuff: kkeyserver, global shortcuts and so on.
> - screensavers
> - netscape plugins
> - the krfb and vnc stuff
> - i surely miss something now...
> on the other hand i worked on the qt/win32 port. i guess i made good
> foundations, on which other people can continue working. i guess the port is
> about half finished now, thanks to chris january and Richard Lärkäng, who
> started working on this some two weeks ago. i hope some other people keep on
> working on this, so that we have a gpled qt on windows native somewhen, so
> that everybody who wants to, can compile stuff with it.
> i got a version of qt 3.1 from the kde-cygwin.sf.net checkout running, into
> which i simply dropped the commercial source files, and made some makefile
> fixes, so that the windows files were used for compiling.
> as i was worried about how to get a usable packet management for the cygwin
> stuff (well, i hope we will soon get some more developers...), i tried to
> understand and use the cygwin packaging system, but imho it is a big mess.
> so i started porting gentoo to cygwin. it works quite nice, and in the next
> two weeks i will get a new server, then i will make it available for
> download. well, one  can discuss about if it makes sense to call that gentoo,
> as it is unly their build system, but it works like in gentoo linux. you want
> mc? do "emerge mc" and it gets the cygwin binaries, in the right order, and
> installs them. i think this is a precondition to use kde on windows: have an
> easy, direct and simple packet management system.
> so that works for me, although it still has some rough edges. i can
> compile qt
> from a cvs checkout from kde-cygwin.sf.net with it, and i am just working on
> kdelibs and kdebase. somehow i have messed arts up...
> what is left to do:
> i guess i will get that running for linuxtag, perhaps i can meet some of the
> core developers, so that they can give me some advices how to get kde really
> platform-independent. if i understood the patches, which benjamin reed did
> for the macos x port of kde, then he is doing about the same, but it seems a
> bit easier on os x, as that is a real unix.
> i will be at the kde meeting in nove hrady, so perhaps we will find some time
> to sort this out in a way, which solves this problem once and forever, like
> the trolls did it in qt. we only have few functionality, which is dependant
> on X, but it is spread all over the sources. so for kde 3.2 that should be
> sorted into some classes, which have different implementations for different
> windowing systems, but which can be used in a portable way.
> then there are some performance issues to fix, like the fork() performance of
> cygwin, which makes kmail unusable, as ralf wrote somewhere.
> when we get that done, the whole kde should be usable on win32 systems as
> well. imagine you can work with kdevelop(gideon), konqueror,
> kmail(kontact/kroupware), koffice, use all the kioslaves, and it would simply
> work.
> then imagine, that when the ct(german computer magazine) will publish their
> next shareware cd, this cd would not only be a knoppix cd as the last one,
> but it would also contain the kde packages for windows.
> double-click on them and they install. that would convince quite some people,
> that kde is better than windows.
> it would be best, if one could compile kde apps on win32 with a gpled qt, and
> one could show how clean the kde sources are to all these visual basic
> programmers...
> but it is still a way to go.
> it would help most, when the other developers would help looking over my
> patches in nove hrady and that could be integrated into the main tree for 3.2
> thanks, Holger
> On Sunday 29 June 2003 17:19, Waldo Bastian wrote:
> > Hash: SHA1
> >
> > On Sunday 29 June 2003 14:16, Vadim Plessky wrote:
> > > On Sunday 29 June 2003 15:02, Waldo Bastian wrote:
> > > |  On Sunday 29 June 2003 12:28, you wrote:
> > >
> > > [...]
> > >
> > > |  > Unfortunately, due to very heavy workload at the work, I even can't
> > > |  > read mailing lists regulary.
> > > |  > May be, some $$$ from my side can help?
> > > |  > I know that it's not good idea to offer money for some Open Source
> > > |  > developement, but don't know wether we have any chance to get Konq
> > > |  > running in Windows otherwise?..
> > > |
> > > |  I think $$$ only works if you can actually hire someone fo some time.
> > > | Note that there is a substantial difference between KHTML and
> > > | Konqueror. You can port KHTML to something other than Qt (Apple did
> > > | just that), but porting the rest of Konqueror will be very hard without
> > > | Qt (Apple wrote the rest of Safari itself)
> > > |
> > > |  Porting over Qt/X11 to build a native Qt/win32 seems to me by far the
> > > | fastest way to get konqy running on Win32. it's a managable job because
> > > | the platform specific parts of Qt are well-defined and not overly
> > > | large. It just requires a few determined people with enough Win32
> > > | experience.
> > >
> > > ok, Waldo, do you know someone who in fact can do the port?
> > >  (and has time to work on this)
> > > Ether with Qt/X11, or Win32 native classes?
> >
> > Maybe Holger or Ralf know someone. Holger started on a port but stopped
> > after getting a commercial Qt license. (See e.g.
> > http://lists.kde.org/?l=kde-cygwin&m=105560257818024&w=2 ) Ralf is very
> > much involved with the cygwin-based version for Win32
> >
> > Cheers,
> > Waldo
> > - --
> > bastian at kde.org -=|[ SuSE, The Linux Desktop Experts ]|=- bastian at suse.com
> > Version: GnuPG v1.0.6 (GNU/Linux)
> > Comment: For info see http://www.gnupg.org
> >
> > iD8DBQE+/wN9N4pvrENfboIRAj3/AJ0ag5iw3C+SarI2bM98FqcbojdI1wCeIQzx
> > uIka6r35ymKn4vyZkyg96bc=
> > =6bKd
> > -----END PGP SIGNATURE-----

More information about the kde-cygwin mailing list