Embedded Linux and KOffice.

I'm on the train back home from ELCE (Embedded Linux Conference Europe) now. It was an interesting conference, not 100%, but quite free software-ish. Met a couple of nice people again: Adriaan (KDE + NLUUG), not quite surprisingly, Armijn (NLUUG + gplviolations.org), but also Will (KDE + Suse) who I've did a bit of networkmanager hacking with in between the talks, Shane (FSFE, FTF hero), Oliver Gravert (Canonical), Valer (NLNet) and others. Feels somewhat like home (which I wouldn't have expected from a mobile / embedded conference in the first place). I gave a talk this morning, which I found quite interesting myself (that's at least one person). I've talked about how to get KDE on a mobile device, starting with an explanation of the software stack and how those parts related to platform independence, moving to technical challenges when turning a desktop system into a mobile system (covering things like screen resolution and pixel density, input mechanism and the more limited hardware you usually find on mobile and embedded devices). People seemed interested although the room was not exactly packed. Also followed talks myself (Will's to begin with, giving me a bit more insight into Solid's networkmanagement stuff, quite useful if you're hacking on that ;-)). Ow, and fixed two layout glitches in battery and networkmanager applet, not bad.

Tomorrow, I'll be on the KDE-road again, going to Berlin to attend the KOffice sprint that's held in the middle of cocktailbars and restaurants (a.k.a. KDAB's office). But why? What do I have to do at a KOffice sprint? Well, one of the topics there is release promotion of KOffice 2.0, which I understand is due in the next months. I think KOffice is in a very interesting situation right now, and it's critical to position it nicely now so it'll have a bright future.
People that know me also might have heard me cursing about OpenOffice. Frankly, I don't see OpenOffice having a bright future. The codebase seems rather convoluted (about as many lines of code as the whole of KDE -- including KOffice) and because of its sheer size OOo is basically unmaintainable. Add to that what looks like a fork by the Novell team decided to publish as Go OOo, and a development model that's pretty much entirely much based on one vendor (SUN or Novell, depending on who you ask ;-)), so community involvement is apparently down to a minimum. With devices becoming smaller and requirements regarding the user interface becoming more central in the Free Desktop ecosystem, I just don't see OpenOffice providing a sensible answer to that.
Enter KOffice. It's free, it's lightweight, it's cross-platform, its codebase apparently is rather clean, in short it could very well fit into that mobile "niche" as use cases are more limited on those devices anyway. So what's needed to get KOffice in shape so that it becomes the serious contender? Frankly, quite some work still, in my opinion. First, it needs to work, a word processor, a spreadsheet application and a presentation program are the least I'd expect. so it makes sense to focus on those apps. Then, support for the most used standards, I know Girish is already working on getting OpenDocument support to work really well (absolutely critical IMO), but at some point, we'll also have to look at support for proprietary formats.

That said, I think there's quite the business case for a manufacturer seeking support for office documents in a certain product. So pitching KOffice to manufacturers in that market makes a lot of sense.

Then, there's of course the platform-indepence thing looming. KOffice, as many other KDE application will be available on Windows and Mac OS as well. How does that impact how KOffice sees and shows itself, and to whom? And how do we do end-user support? We certainly wouldn't like to burn credits by people that run into simple problems but just aren't able to find a solution to their problem or support in general (be it community support or commercial support).

Another interesting angle is the "OOo reinvents MS Office so we don't have to angle". KOffice's current user base is pretty small, so in turn, there are very little people who will complain when KOffice goes new ways (like Plasma did in KDE4), so I'd expect much less pushback when actually fixing the "Office UI", making it a sensible interface that's is a pleasure to use instead of something that drives me nuts the instant it starts (which is not really an instant in fact ;-)).

[ Thu, 06 Nov 2008 18:55:58 +0100 ] permanent link

This weblog does currently not offer the option to comment. I would be happy to receive an email with your thoughts.

Weblog Archive