Hi,

It's been almost 6 years now that I use GNU/Linux and I am very happy with it. One of the sad think that I find out is that sometimes instead of unite on one project to make it become better, teams prefer to develop in there own corner (which by the way could be a good alternative for some project). I think for example that on the office side we should try to unite our force and stop having Openoffice.org on one side Abiword on the other and Koffice in between. All of them are great and offer some nice features but couldn't we just put everything together to have the best office suite available. I think that we may have to review the way that we are developing our software. Instead of just focusing on which desktop environment we want it to work and to avoid falling in the eternal toolkit war GTK vs Qt. We may should develop the core component of the office suite in an abstract language and then just let the look & feel handled by what ever graphical library we want (it's maybe already the case. I don't have much knowledge on how every office suite are developed). I think that with the opportunity offered by Libreoffice it's maybe time to try to join our effort for a better united office suite that will be beneficial for everybody.

That's just my 2 cents issue from my observation of the Free software community. Unfortunately I don't really know from where we need to start to make this possible. I would like to do something about it, but don't really know what so that's why I decide to wrote this topic to try to have more idea or try to unite some interested people in such approach. I really think that we really could achieve something in GNU/Linux world for many software if we try to take all the great idea available and combine them to make it better. I know that may seems to idealistic but if we look at all what the free software community achieved so far, nothing should be impossible

Ps : Sorry if my English is not good enough, I hope that it will help you to understand what I mean.

It is always a tricky balance; the difficulty with having one of something is that it will never meet everyone's needs or preferences. The difficulty with more than one is that you risk duplication. I think the co-operation between KDE and Gnome is the right way to go - find things which everyone needs and co-operate on those but then offer different things to meet different needs.

BTW I use Kwrite, Kate, KWord, LyX and OpenOffice, each for different things; I couldn't really manage with one (though at a pinch I could manage with three). So I am pleased that there are alternatives available.

If we create something that is flexible enough for a specific kind of software. I think that we can manage to cover most of the needs of the users. But for that we need to have a common base where we can build plug-ins or module to add functionality.

In your example you are using different kind of software : Kwrite (simple notepad), Kate (a powerful code editor)and Lyx (a specific tool for doing LaTeX). The only tool that could be comparable are the office suite. We are all attach to choice in the free software community and it's a really good think that able us to not be constraint and drop our liberties. Basically what I mean by unity is to work all together to build the basic element like we are doing for the Linux kernel. The same technique should be applied for the major application such as office suite, pdf readers, image manipulation, music player and so on.

We need A basic ground for all this projects and then we can build upon what ever we would like to meet our needs (desktop integration, lightness, customisation...). Maybe a way to achieve that is to offer for every program a package with the core functionality then we can just add what ever front end we want for it. I know that such approach have been done for some core functionality but the whole software should follow this model in order to stop the futile war for Graphic Library and the multiplication of software doing just one part of the job.

This unlikely to happen in the foreseeable future.

There are several reasons:
There is a well established codebase in the applications. Basing them on a common core would mean that you have to replace several hundred thousand lines of code. That is code that has been develop for more than ten years and would take years to be replaced.
Of course there might be some smaller parts that could be shared, but most of these are usually too small to matter.

The office suites are based on different toolkits (even in the lowest level), different programming languages and different licenses. They even follow different development philosophies.

It has been tried before and didn't work out. There was libwv that was shared between AbiWord and KWord, but the development split again.

There isn't an office suite that I know that combines the features of Kwrite, Kate and LyX; the last program I came across that combined code editing with high quality output was Newword in the 1980s - at a much lower level of course than we are accustomed to today.

The problem is partly the obsession with WYSIWIG which is so resource hungry; I don't want to open a monstrosity like MS Word or OpenOffice to do a short note. I want something I can open - write a note - and close almost as fast as I can hit the keys.

The joy of LyX is that I can get on with the job of writing or scrolling through a document without having to wait for the screen to update every few seconds and I can work on documents which bring MS Word and OpenOffice to a halt because of their length.

Your idea isn't a new one. It fails every time because you're not going to get everyone to agree. Feature-completeness and interoperability matter more to people than an idealistic framework. Why not use an OS like Plan 9? It is so much more ideal than a clunky old Unixish OS like Linux.

Choice is important too. One implementation of anything isn't going to satisfy everyone. Relying upon one of anything is dangerous. It isn't time to unite, it's time to decentralize and distribute. This is why it can be so hard to run KDE apps inside Gnome, or stand-aloneish applications as replacements for various KDE components. Everything is TOO centralized, which generally goes against the unix philosophy. Change anything and it breaks.

Hey,

from a technical perspective this idea makes a lot of sense. And to give a bit of background what I mean with that I have to explain what makes up such an application. The boring parts is what a Windows APIs deliver. The buttons, the text rendering etc. If you would use Windows APIs all you'd have to write is the user interaction.
None of the applications you described use the Windows APIs. OpenOffice wrote a whole set of libraries for text rendering, for buttons and tabs etc for themselves.
KOffice uses Qt for all of this.
AbiWord uses GTK where possible.

So, what I would suggest for uniting is everyone to use the same stuff to build on top of. Naturally I'm going to suggest to use Qt because thats what I believe in. Using the Windows APIs instead sound wrong, I want it to run on other platforms too. Using the OpenOffice uno stuff sounds wrong too. The main reason for that is that the OOo community is there to write an application, the stuff I described above comes in second. I'd rather use the toolkit from a dedicated company that has been in business for just as long and has been focusing on only the toolkit.

So, for uniting I think we are at the point where it takes too much work to change. I don't see OOo start to use Qt, same for AbiWord. The reason is given above, it would take too much effort.

I wanted to look at this suggestion from another perspective too; the perspective to just drop KOffice and move all the developers over to libreOffice. Completely ignoring the human aspect of it, I think its worth to have KOffice in place for several reasons.

KOffice is one of the main ODF implementations next to OOo. Having more than one implementation is very valuable to make sure human mistakes in implementation get caught. In other words; it makes the standard better.

KOffice is not aiming to be just another office app. We have several already and it makes little sense to chase after a 20 year old application and idea. KOffice has a whole set of new ideas and a new approach to using an office suite. Its much more extensible and thus more easy to fine tune in business that would otherwise have to get a whole application build for them from scratch. Think about the music shape; a small component that essentially turns KWord into a music editor and printing application. Thats extremely powerful!