[prev in list] [next in list] [prev in thread] [next in thread] 

List:       koffice
Subject:    Re: KDE version of OpenOffice?
From:       Philipp =?iso-8859-1?q?M=FCller?= <philipp.mueller () gmx ! de>
Date:       2002-11-01 6:34:49
[Download RAW message or body]

Hello Piotr,

Am Donnerstag, 31. Oktober 2002 15:46 schrieb Piotr Gawrysiak:
> I hope that this post is not too off topic and that it is not too naïve
> etc. - if it is, please excuse my stupidity :-)

> Have anyone considered doing something that could be called "KDE port" of
> OpenOffice? 

Don't asume, that I or the other developers haven't thought at least even a 
bit of it, but it is rather impossible to make a simple "port".
OO or StarOffice uses it's own widgt set which is in principle even a complete 
own dektop environment like GNOME or KDE.
So by trying to port OO, you will need to change everything. So the only thing 
what would remain would be some data and functional logic. Realy not much at 
all and in principle the same as if we would "simply" look into the code of 
OO and take something into KO.

Nobody stops anybody from trying it, but nobody realy started it as well as 
trying to port OO to GNOME. It's simply a very hard and maybe wrong approach.

> Of course it is possible now to run OpenOffice alongside with
> KDE but it is not well (or at all) integrated with the rest of the
> environment - it has its own graphical widgets, own font handling, own -
> well pretty everything (see note 1).

As I said it has rather it's own complete desktop, not only widgets.
That's why even simple integration in other desktops is rather hard with OO.

> So it looks a bit out of place on KDE
> desktop. And also behaves a bit strange - what is probably most annoying is
> that more sophisticated copy/paste/embed operations are of course not
> possible. And it would be so natural to be able to take, say, KFormula and
> embed in OpenOffice document :-( (see note 2)

Additional embedding is in OO a different technology in comparision to KO.

> I do not know how much effort would be needed and if it would doable at all
> but - what about taking OpenOffice code and porting it to Qt/KDE? So that
> the resulting application would be compatible with other OpenOffice
> versions, for example Win32 (see note 3), but would use KDE libraries,
> integrate with the environment (toolbars, file open dialogs, printing
> system, KParts - for example embedding KSpread in OOo Writer etc. etc.).
> Maybe this could also make it a bit faster and smaller.

No, not easier or faster. This IS the hard and long part.

> I am not sure about this - but wouldn't it be a bit similar to that what
> OpenOffice Mac port is finally going to be?
> Currently it looks (i.e. the
> interface) just as other ports but I think that there are plans to make it
> to Aqua. Or am I wrong?

I don't know much about the Mac port, but I assume they don't want to change 
the internals, but simply try to adjust the OO widget design to the aqua 
design. This is not realy integration, it's only mimic AQUA.
But here I can be wrong of course.

> Of course there is KOffice. But in its current state (especially
> considering lack of developers and - what is probably much more important
> and what might be the reason for it - lack of marketing and media
> attention) it somehow does not seem promising (see note 4) :( 

You are right, lack of developers is a problem (but a problem that every Open 
Soiftware project has - tell me one that has enough).
With marketing you are rather wrong. I think we are getting more of media 
attention than what the current state of KO is worth it.

Okay, we don't have a native Win32 or a native Apple port, but both I'm not 
very interested in. If others like to, fine with me, but not my personal goal 
(I'm one of the KSpread developers). There are ports to both environments, 
but they are working in an "emulation" mode (the XServer is emulated and 
slows the stuff down).

> Even outside
> of Linux community (well, not that I do not use Linux, but I am still
> rather Windows-only specialist :( it seems that it is possible (still only
> just possible, but it is something) that OpenOffice might be a real
> competitor to Microsoft Office and even replace it in some quite distant
> future.

Fine, I love that there is OO. If OO will have more market share, then MS 
looses dominance and users learn not to stick to things MS is dictating.
OO is good for the whole Open/Free software world, as at the end MS and the 
users will be forced to look for open standards (e.g. file formats).

This doesn't mean that we need to now stick to OO at all. OO has a rich 
functionality but not all we want in detail. And it never hurts to have other 
solutions as well. Monolitic is a bad thing.
The only difference, why it looks so promising to get away from MS with OO at 
the moment is simply that OO was StarOfffice for years and are in power man 
years ahead of other projects. So in the current state, the only real 
competitor to MS is OO. But this shouldn't be the end and in some years we 
will be fully compatible as well.
And just to say: KO already provides a lot of feature which a big user base 
needs. We are not at the end of KO and I would rather say we are just at the 
beginning.

> That does not mean however that this imaginary KDE OpenOffice version and
> Koffice could not coexist, or even benefit one from each other.

As I said, I doubt there will be a KDE OO version.

> What do you think? Does it all make sense or am I only dreaming? Is it
> worthwhile to start thinking about such project (gathering support,
> developers, and maybe - even funds) or is it just stupid?

Wortwhile it is yes, as worthwile as SUN thinks about a GNOME port (I asume 
they do as they said about it officially), but I haven't seen any line of 
code yet from anybody.

> note 1 - I think one of the Koffice developers was investigating Ooo source
> code trying to assess if the import/export functions could be reused in
> Koffice and had some remarks on this topic?

I think most of use already looked into the code of OO, only to see that OO is 
in principle a big monolithic thing, where it is hard to take some parts over 
to KO. That doesn't mean that we don't learn from OO, it just means it's 
harder than on the first thought.

> note 2 - Win32 version of OpenOffice integrates much better with Windows
> environment. For example OLE embedding is possible, so one can take say -
> Microsoft Equation object and embed it in OpenOffice writer document - it
> is not perfect, but more or less works.

Yes, that's because they use OLE as embedding technology. I cannot say what's 
better, but I trust the people who defined KParts when they said OLE has some 
fundumental design mistakes and even QT stopped there implementation just 
before the release of QT3 (still they want to include something in later 
versions).

> note 3 - by compatibility I understand document file compatibility, general
> properties of the user interface, and also - what seems to be very
> important - the possibility of incorporating future OpenOffice source code
> improvements/changes into this KDE version of OpenOffice.

Document file compatibility is planed already, file compatibilty has nothing 
todo with the technology behind an application (beside of the simple fact 
that if one app supports tables, the other one must do it also to use the 
file correctly). The other stuff see above.

> note 4 - which is a shame, because KOffice has lots of potential. However -
> this is just as with KDE vs GNOME wars - does Open Source community have
> enough resources to push two similar projects at the same time now?

No, not a shame and:

    There is no war between GNOME and KDE, not at all!

The war is only a marketing wording from newspapers, so they can sell their 
stories better.
Between GNOME and KDE is competion not war!

Both environments are only doing the same, as we find it with nearly all free 
software projects like OS's (FreeBSD - Linux), printer servers (lpd - CUPS), 
web servers (Apache and all the tiny ones), proxies, browsers, mail clients, 
...

No problem from my side here, monolitic structures are bad bad bad bad bad...

The choice is the good thing. The choice forces improvements. The choice tells 
the users and programmers, there are other ways of doing things. The choice 
drives the future!

Haven't you already learned how much MS has slowed down user friendlyness only 
because the user didn't know it can be done differently and MS didn't care?

I asume you mean the same here as you want to have the choice against MS and 
OO is not yet 100% competitive (maybe 95% already - enogh difference for 
some).
But that doesn't mean replace monolitic MS with monolitic OO. That also 
doesn't mean (and this is the important part), that we are at the end. It 
only means, today you have nearly 1 choice, and trust me, in future you will 
have much more choice when the other projects aren't stoped.

> One example - I think that Lycoris now has the "Office" component - which is
> somewhat modified OpenOffice and not KOffice. And this is KDE only
> system...

Somehwat modifies means here, replacing the words of the programs (something 
like changing 10 lines of codes) and remove stuff they don't need (like 
translations) and not giving the users to upload any bugfixes from OO anymore 
and putting in some shortcut icons.

Final words:
Just be patient. Linux wasn't born in 2 years, it took 10 years to become 
competitive. KOffice is now here for some years, but we are still in the 
starting point. And if we don't get more developers now, okay then it will 
take some years more, but at the end we are there.
And Office work didn't change in the last 10 years fundamentally (I can still 
do nearly all jobs with MS software 10 years old, only some small things are 
missing). So don't expect that we will always be behind the task, the task is 
defined and it will be reached and when not tomorrow then in the time after 
tomorrow.
I will not stop on working on KOffice, I have the patiency.

Philipp

PS: I'm not talking for KOffice in general, I'm just giving my impression and 
view of things. Other developers may see it differently.
____________________________________
koffice mailing list
koffice@mail.kde.org
To unsubscribe please visit:
http://mail.kde.org/mailman/listinfo/koffice
[prev in list] [next in list] [prev in thread] [next in thread]