KOffice 2 Alpha 8 Reviewed

KOffice, like most KDE applications, has the unique selling point that the same codebase can be compiled and run on various platforms. The latest KOffice alpha release has attracted extra attention since it ships Windows and Macintosh binaries. This naturally means a much larger audience will be able to try it out and kick the tires. Techworld has an article up where they give a good overview and plenty of screenshots. Have you already tried out a KOffice alpha release?


Some clarification from my side on the article - http://blog.forwardbias.in/2008/07/koffice-alpha-8.html

By girish at Fri, 2008/07/04 - 5:00am

Guys, we're looking for volunteers to make screenshots for 'visual changelogs' of each release.

More info at http://www.kdedevelopers.org/node/3545

By js at Fri, 2008/07/04 - 5:00am

This sounds great! We need screenshots for us who dont have space/time/whatever to download and install KDE4daily/SVN etc. I'm so waiting that kword has space around sheet, so it would look more like OO.o or MS Office. I just cant stand how the sheets are together, and I want to get just verification about this usability thing! :-)

By Thomas.S at Fri, 2008/07/04 - 5:00am

> I'm so waiting that kword has space around sheet

Hmm? KWord has had that for about 3 years now...

By Thomas Zander at Fri, 2008/07/04 - 5:00am

It has not have space _around_ sheets, only space on it's sides but not between sheets, what is important. Or then I have not found option to get space between sheets. :-/

By Thomas.S at Fri, 2008/07/04 - 5:00am

Ah, got it. Thanks for the clarification :)

KWord can do that too in the 2.0 series. See http://www.koffice.org/kword/pics/200606-KWord-PageSpread.png

By Thomas Zander at Fri, 2008/07/04 - 5:00am

Ah... good to know. I hope that space can be adjusted to be bigger than just a few millimeters, like between 1-15mm what can be set from options. Then it would give better usability to sheet view, so user can see more easily the difference of two sheet and how it would look on printed.

By Thomas.S at Sat, 2008/07/05 - 5:00am

I like the new interface in KWord 2. I think that it could help encourage using semantic markup (head 1, body text, ...) rather than low level visual formatting. Some suggestions to further encourage users to do the right thing:

1) Put the semantic formatting options in more prominent positions, on top etc. For instance, the "paragraph" tool tab (which contains semantic formatting) could be the first one, rather than "style", which is low-level.

2) Include semantic character styles, like "emphasized", "quote". Maybe put them on top of the style tool tab, then put the low-level formatting (bold, italic) below a separator line?

3) Include a "body text" paragraph style in the standard template. This should include formatting to separate paragraphs (indentation or vertical white space depending on the template). Set this style to be the "next" style for all the paragraph styles, except "standard". Maybe rename the "standard" paragraph style to "unformatted" or something that does not encourage its use as the standard style.

4) Include a style browser that lets users browse, and customize, various templates for the standard formatting styles, with a live preview.

I have some more suggestions, but I'll stop here for now.

Thanks for some great applications!

By Martin at Sat, 2008/07/05 - 5:00am

Those are some good ideas, right there.

By Micah at Mon, 2008/07/14 - 5:00am

You do understand what that word means right?

By fffadas at Sat, 2008/07/05 - 5:00am

I'm pretty sure he knows what unique means and that he knows what he is talking about. http://www.merriam-webster.com/dictionary/unique:

1: being the only one
2 a: being without a like or equal
2 b: distinctively characteristic
3: unusual


Oh I know you want to argue that KOffice is not the only office suite that can be compiled natively from one code base on all available platforms. So please show us one known office suite that has the same feature (hint: It can't be OpenOffice - you cannot compile it as a native app e.g. on Mac from the normal code base).

By Arnomane at Sat, 2008/07/05 - 5:00am

> hint: It can't be OpenOffice - you cannot compile it as a native app e.g. on Mac from the normal code base

I think this will be the case for the next major release, or am I mistaken? And KOffice 2.0-final isn't released yet, either.

By panzi at Sat, 2008/07/05 - 5:00am

A few screenshots and a single notion that KOffice works slowly.

Is it a review?

By Artem S. Tashkinov at Sat, 2008/07/05 - 5:00am

I'd like to see Trolltech branch out and produce some filters for ODF and Microsoft Office within Qt that all developers and applications can access in a usable fashion. Not only would it be a big step forward for Qt and KDE applications, I'm pretty sure a lot of Qt customers would be happy as well.

By segedunum at Sat, 2008/07/05 - 5:00am

Couldn't KDE just make some as part of KDELibs. A QDoc (or something like that) would be cool (and make some Qt customers happy), but I think KDoc in KDELibs would work just as well.

By Riddle at Sat, 2008/07/05 - 5:00am

Perhaps something like a OpenDocument library?
Take a look here: http://dot.kde.org/1177773146/

By Morty at Sat, 2008/07/05 - 5:00am

I thought and ODF library was happening. I have seen it discussed quite a bit.

What I'm curious about is the MS support. Some people claim it isn't important because people SHOULD use open standards. I'm all for using ODF for my files. But what about the litany of files I need to open at work in a Microsoft shop? What about converting existing users' documents, even if they going to use ODF in the future?

Honestly, until you can reasonably import and export the MS binary formats, you can't expect most users to switch. I use OpenOffice when I can, and people brag up their excellent MS import/export capability, which sadly isn't all that great.

I've read why the MS formats are as complicated as they are. I understand reverse-engineering a closed, binary format with no help isn't easy.

What I don't necessary understand is why any work on deciphering MS formats isn't put into a shared library that everyone can use. I've seen it asked if KOffice can borrow the superior implementation in this regard from the OOo sources, and each time I've seen a resounding no answered, but I'm not sure why that is.

Can someone please explain why that is?

By T. J. Brumfield at Sun, 2008/07/06 - 5:00am

There is quit a lot of work done to get import-filters for MSOffice done and there is even a gsoc that deals with that topic. That export has lesser priority is imho even logical taken into account that MSOffice 2007 will come with an ODF-import filter. Reusing the OO.org import-/exportfilter is out of scope, just look at the code and you will see what I mean.

By Sebastian Sauer at Sun, 2008/07/06 - 5:00am

It's difficult because MS Word files are basically dumps of Word's internal data structures, and OOffice (which inherited their import filters from StarOffice) basically maps them directly to OOfice internal data structures. Trying to adapt this to be useful for KOffice is possibly even more difficult than writing a new filter from scratch (and this is missing manpower).

There have been several attempts to create a shared library for importing MS stuff, there was one called wvware, which was apparently shared by Abiword, Gnumeric, and the like (i'm pulling this from memory, forgive me if I'm mistaken here).

Getting the basic formatting and fonts right is easy. Getting complex OLE structures and macros out is very very difficult.

By KOffice fan at Sun, 2008/07/06 - 5:00am

>there was one called wvware, which was apparently shared by Abiword

And also KWord. KWord still uses the descendant, wvware2.

By Morty at Sun, 2008/07/06 - 5:00am

If I remember correctly, work on wvware2 was even spearheaded by KOffice developer Werner Trobin.

By Kevin Krammer at Sun, 2008/07/06 - 5:00am

> I thought and ODF library was happening. I have seen it discussed quite a bit.

My guess is that the current priorities are on finishing and stabilizing the current implementation and have such a library as a goal post KOffice 2.0 release.

It is important to understand that moving a project-internal library to a general library module (e.g. kdelibs) is a lot of work, since one has to remove all the project specific code, adapt the project's code to cope with these changes and put lot of work in API clean up because of the more strict change restrictions of library modules.

I am very confident that everyone involved understands the importance of widespread availability of ODF capabilities, so it can be used as an almost universal format for import and export not just for files but also for clipboard or drag&drop operations.

> What I don't necessary understand is why any work on deciphering MS formats isn't put into a shared library that everyone can use.

As someone else already wrote, the Free Software office developers (e.g. KOffice, Abiword) have been working on such libraries for quite some time.
It is a bit unfortunate that the formerly proprietary code inherited by the OpenOffice.org project wasn't available in such a form and thus couldn't be contributed to a common effort.

At some point it might be possible for the OOo project to leverage the existance of the shared format ODF to create a stand-alone converter which could then be used by other Free Software projects to tranform legacy formats into ODF and read this.

By Kevin Krammer at Sun, 2008/07/06 - 5:00am

"The latest KOffice alpha release has attracted extra attention since it ships Windows and Macintosh binaries. This naturally means a much larger audience will be able to try it out and kick the tires."

By Unga at Sun, 2008/07/06 - 5:00am

"The latest KOffice alpha release has attracted extra attention since it ships Windows and Macintosh binaries. This naturally means a much larger audience will be able to try it out and kick the tires."

Its good when Koffice runs successfully on Windows and Mac, the so called "larger audience" has no reason to switch to more secure, fundamentally correct Unix, Linux, FreeBSD, etc. No wonder Microsoft offer full support for Blender developers to get the Blender run successfully on Windows.

By Unga at Sun, 2008/07/06 - 5:00am

You just said why they would switch to *nix: because it is more secure and fundamentally correct.

By Michael "Unga i... at Sun, 2008/07/06 - 5:00am

Don't forget, once they're using cross platform software it becomes easier and easier to switch. You may not dump your current OS, but in the future your next system could be one running all free software (few people manually install windows themselves, expecting them to install a Linux distro is IMO not realistic).

*Nix based systems are slowly becoming more common and popular (as options from manufacturers) on non-server systems, so it's becoming more and more realistic to use this strategy. Does anyone really expect people (especially non-techies) to manually install a different *operating system* just so they can use an application they've never used or possibly even seen or heard of before, giving up most all software they're used to using?

By Kitsune at Mon, 2008/07/07 - 5:00am

Hello Guys,
While I admire the energy/effort being placed in this product I would think that with other good/existing open source office suits (OpenOffice comes to mind) that it would be beneficial for the KDE team to focus their energy on putting together a spectacular desktop and use "all their members/developers" to ensure that KDE4.1 is built as quickly (and stable) as possible.

I doubt that Koffice will change the way people use office suites and doubt even more that it will have as big an impact/adoption as openoffice 3.0.

While I admire the tenacity (i really thought the project would lose steam) I think someone should suggest that a moratorium be imposed on extraneous elements of KDE until 4.1 is complete and stable (enough).
My question is simply this, does anyone believe Koffice really is necessary (now) and whether it would not be better that some (smaller) effort is spent to integrate/extend OpenOffice (if even a custom version of OO)?

I like what i see with Kword,Kspread and Kpresenter and I am sure it holds up well with regards to memory use and hence would be practical on small (cheap) laptops. However, in an era of Google docs, 280slides.com [http://280slides.com/Editor/ ] and ever more powerful RIAs (using AIR and JavaFX) I see most of the current desktop apps being made irrelevant in the next few years (especially when ODF and OOXML take root).So the question is how much return on investment (time and energy) will be obtained from devoting so much time to this suite (possibly a silly statement for an open source project).
Therefore, I would think that a well designed and scalable desktop with properly defined interfaces would be the better place to put focus.

-Kurt [BP]
Also there is a cool graphics app written for the AIR runtime and funny enough the performance is already quite good and with Adobe launching Acrobat.com [http://www.adobe.com/acom/] I think we will soon see internet delivered applications comparable to full desktop suites very soon.

By anon at Sun, 2008/07/06 - 5:00am

This is assuming three things:
1) That all of the developers care about the desktop: some developers are on Windows and can't/don't use the KDE desktop. It is unreasonable to ask them to develop something they can't test!
2) That all developers want to and can develop the desktop: some developers are not familiar with and don't want to become familiar with the desktop code base. Others simply are better at developing office suites than developing desktops. Again, it would be unreasonable to ask them to develop something they don't want to develop.
3) That the office suite does nothing for the desktop: at the very least, KOffice helps test/refine the platform as used in different situations other than the desktop. At the very most, the desktop may make use of platforms developed as part of KOffice: for example, Kross started as being developed in KOffice. Other platforms, such as Flake and Pigment, may eventually be moved into the KDE libraries as well.

By Michael "Replyi... at Sun, 2008/07/06 - 5:00am

1) Free software developers code for what they want to because they want to. Only a small percentage are paid for it. What this means is that you can't magically reassign a team working on one project to another project. They're working in their spare time, usually from 10-20/hrs/week *on top* of whatever else they're going (ie full-time job + family + life).

If you think you're going to tell one group of people to go code on whatever KMagicPony of the week there is, you're crazy.

2) I think that KOffice actually existed before OpenOffice. So maybe all the OpenOffice devs should join us? ;)

3) People seem to like saying that choice is good, so I might as well repeat that. Although I'd say that KOffice is probably good for KDE because of its tight integration, etc. If we really wanted to, we could probably outsource most of the desktop to elsewhere and other projects, but then we cease to be KDE.

By A. L. Spehr at Sun, 2008/07/06 - 5:00am

Thank you, I wholeheartedly agree with this sentiment.

Good on the KOffice team, I for one, will use it.

(I know people support Open Office due to their support of legacy formats, and the "only game in town" mentality, but viewed in its own merits I feel it no amount of "push" is gonna make a nimble nymph out of it)

The original question really grates my carrot, when people like this attain positions of power, we as open source users/devs, are porked

By Donald Carr at Mon, 2008/07/07 - 5:00am

> Adobe

Those Adobe that fails since years to compile a 64bit Flash? :)

By Sebastian Sauer at Sun, 2008/07/06 - 5:00am

AIR is an interesting example. it uses WebKit. perhaps you would've said the same thing about khtml 5 years ago as you just did about koffice. perhaps you even did. but had khtml not been around there also wouldn't be a WebKit and maybe AIR wouldn't've even found its feet as a result.

so why koffice?

because it was there first (before any other open source office suite), because the developers want to, because there's nothing like some of the apps in the foss world (kexi and krita in particular) and because it runs well on very modest hardware (when's the last time you saw OOo on a handheld device?).

who knows what koffice will become in the future.

and to be honest, unless you or i are working on it, it's none of our business to question it either because it's not our investment of time, energy or money.

finally, the "put more people on it and it will go faster" thing is a complete myth, just as the "if you make people not work on Y then they'll naturally work on X" idea is a myth.

By Aaron Seigo at Sun, 2008/07/06 - 5:00am

Right on Aaron! And let's not forget the demise of rationality in zero sum economics. People assume a developer working on project B takes a developer away from project A. Why should Linus Torvolds have continued working on his kernel after he met RMS instead of going to work on the herd kernel? In fact the more developers work on different projects, the more different projects developers start. Maybe some ought to see if there is one in line with their thinking before starting, but it is absurd to think developers could be shuffled off to other projects by some centralized big brother. Free software is about freedom and choice. In fact the idea of terminating a project that a number of users love because it has failed to achieve a majority of users is really rather fascist. What standard or timeline should be used? Do we go for a monolithic program and eliminte choice? I thought that is why people like me were using GNU/Linux...

Finally, users refer to developers as if they were born developers, when in fact most developers learned to become developers to write code on projects they were interested in. That's how and why I learned C++. I was a user swept into an infant project when there were many alternative programs to Quanta out there. I wanted a program more to my liking and as a result my program became the most popular in it's class... like other KDE programs.

A better question than why someone is doing something they are contributing to the community is what are you doing and contributing? I view the contribution of any program that any user downloads and enjoys at a substantially higher value than any opinion that does not produce a real benefit to users, thus to the community. It is in our nature as humans to mistake intent for action and think we are doing something when we merely want to have some impact. To be clear, without people giving of themselves to make the software happen the community disolves. So I think the best rule is to never imply someone should not contribute in a given place, because there is no zero sum distribution of programming and they probably won't jump ship from their program. I wouldn't.

What is necessary is development on projects. Suggesting otherwise is ludicrous.

By Eric Laffoon at Sun, 2008/07/06 - 5:00am

and right Eric!

Guess myself would just have left linux and went on to another system if KDE wasn't there in the first place and caught my attention :)

By Sebastian Sauer at Tue, 2008/07/08 - 5:00am

I'd rather have KDE 3.5 with a excellent KOffice. KOffice is the best office suite. OO is so slow and arcane. It's like Cakewalk. You've got layers of old code. KOffice is very fast. I like fast simple software.

Besides, KDE is not communist. People do what they want. ;)

By winter at Sun, 2008/07/06 - 5:00am

I agree.
KOffice is fast, it is simple (from a user's point of view) and it is just doing it's job. What more can one want?

And for the really nasty things there is Latex ;-)

By captain obvious at Mon, 2008/07/07 - 5:00am

Does that make KDE anarchist then? lol

By NuclearPeon at Wed, 2008/07/09 - 5:00am

Seeing apps like Krita there is no doubt the answer is: YES.

By Iuri Fiedoruk at Mon, 2008/07/07 - 5:00am

Because one size or fashion does not fit all.

Why do peaople use MS Works when there is MS Office?
Why do people buy Mac when there is DELL?
Why do people use Gmail when there is Hotmail?


KOffice is not an exact copy of OpenOffice. They seem quite different.

By Axel at Wed, 2008/07/09 - 5:00am

First of all, KDE applications are meant to function in a KDE environment seamlessly, something that couldn't be achieved in Open Office or other Office Suites.

I personally do not like open office compared to KOffice because the startup times are longer, although this issue is trivial. That aside, KOffice just 'feels' better, less out of place, plus it has that famous KDE functionality going in its favour.

I really dislike Microsoft Office, so having KOffice look different is also a plus. I would prefer that KOffice act more like Kile in regards to formatting. (Kile is a Latex editor, so formatting doesn't show up until rendered)

For my last point, competition is good for the market and consumers. Who said that the only contenders could be Microsoft Office and OpenOffice? That's a rather narrow way of thinking if you ask me.

Plus I think the Open Office windows port looks horribly ugly. But that's not relevant here.

I work with adobe almost every day, and I'm really unimpressed with the GUI layouts especially in reader and acrobat.

By NuclearPeon at Wed, 2008/07/09 - 5:00am

to me the unique selling points of KOffice are a quality FOSS database GUI (kexi), a really cool imaging app (kexi), that it's small and fast and integrates nicely with the rest of my dekstop.

but that's just me.. =) congrats to the koffice team on spreading your wings further.

By Aaron Seigo at Sun, 2008/07/06 - 5:00am

Personally I love Krita. It deserves more attention, especially in light that it's much better than gimp :)

By Mark Kretschmann at Sun, 2008/07/06 - 5:00am

Seconded. I'm eagerly waiting for the next iteration of Krita, since I liked the first one quite a lot. (Plus, it's much better integrated into KDE than the GIMP)

By Luca Beltrame at Mon, 2008/07/07 - 5:00am

At the Linux Foundation Summit, an Access-replacement was given as one of the reasons one of the companies didn't switch some of their desktops to Linux. So further Kexi development and promotion is quite important.

Not that Kexi is any use to me personally really. ;) But Krita is a fun app for playing with my small drawing tablet. I'll probably try out Krita 2.0 once it hits beta or so. I look forward to the more natural brushes.

By Ian Monroe at Mon, 2008/07/07 - 5:00am

Great to see that KOffice progresses well (especially on Windows/Mac OS X). With its great technology (clean codebase, well integrated/flake) and its completeness (traditional office apps + krita + karbon +...) it has the possibility to become a major office suite.

This will for sure require some more helping hands. I am very optimistic that this is achievable due to the attractiveness of the KOffice platform (well designed and inter-platform). A bit more promotion/hype (e.g. something like Troys' "the road to KDE 4") could surely help to spread the word, too.

Well done KOffice team and keep up the fabulous work!

By MK at Sun, 2008/07/06 - 5:00am

just to say thank you for your work on kplato, and a personal thank for Danders on his work on EV analysis,

By mimoune djouallah at Sun, 2008/07/06 - 5:00am

Couldn't agree more on this. KPlato is an excellent project planner that you can't find in OpenOffice.org and I also have to say thank you very much to the authors of KPlato.

By Jure Repinc at Mon, 2008/07/07 - 5:00am

A guy wants to draw, so he takes over Krita, builds a team. We get a math library out of it. They come up with stuff no one else has, like mixing colors.

Wouldn't it be neat to do database stuff? Someone starts writing Kexi. Years, yes years later, we have something pretty cool.

And so on. Every app is a story, a reflection of someone, or of many.

The hubris is awe inspiring. Their persistence is humbling.

I would use these apps for that reason alone.


By Derek Kite at Mon, 2008/07/07 - 5:00am


By taj at Mon, 2008/07/07 - 5:00am

Nice post!

By borker at Tue, 2008/07/08 - 5:00am