ESR's wife Catherine has gotten herself entangled in the issue of usability for open source office apps. Sure, as hackers we like to pat ourselves on the back about how far we've come along over the last few years. PostgreSQL and MySQL are quite suitable replacements for commercial database engines in many situations. Linux is getting an increasing share of shelf space in the glass house. And so on.
But the desktop software situation has always been sadly lagging behind. Following a disappointing experience with KPresenter, Catherine is giving voice to the secret that dare not speak it's name: open source office apps aren't really that good. There, I've said it. No more dancing around the fragile egos that have spent countless years developing these apps. In the end, wizzy toolbars and constantly evolving file formats don't amount to a hill of beans. Catherine is among those that are quite vocally pointing out that if the software isn't easy to use (and can't use the formats du jour -- like .doc files) then the open source office apps are suitable only for the most ideologic fringe of geekdom.
What's the source of that problem? One of the most common issues in the industry: developers and users speak two different languages and rarely see eye-to-eye.
Thankfully, Cathy is interested in bringing these issues to the fore and getting them addressed. Perhaps then we'll see realistic open source office apps that have a chance of being useful beyond the lunatic fringe. It seems like the impediments are quite substantial barriers to adoption, but relatively minor refactorings of what's already been implemented.
Re:Hard work
ziggy on 2002-03-16T14:54:53
People vastly underestimate the amount of work that has to go into something like an office application.No doubt about that, and the Star/OpenOffice folks, KDE/KOffice folks and various Gnome folks have done an huge amount of work over the last few years. I daresay the amount of time (in man-years) that's gone into these projects far surpasses the amount of effort gone into the Linux kernel by an order of magnitude or more. It's also a much harder problem to write software that DWIMs, rather than device drivers and filesystems that do what they should.
I don't think I'm underestimating the effort though. I applaud what Cathy Raymond has done by being the frustrated yet helpful end user for KPresenter. It shows that the KPresenter team has done a lot of great work over the years. But if they don't account for the needs and requirements of the end user, they'll have a damn fine 80% solution that never works for more than 20% of the audience (more realistically, 2% of the audience).
As programmers, many of us have waived these issues away as simply "window frosting" of some sort. But in this type of development it's the window frosting that matters most: overall usability, consistency, interoperability and doing what the end user expects after years of working with MSOffice. This is an area where superior technology doesn't amount to a hill of beans if it doesn't meet expectations at the end. I really think that the KDE folks are truly almost there, but need a little more input from what the end users out there really want. Perhaps at that point we'll see a legimate (superior!) alternative to MSOffice on freenix desktops.
Re:Hard work
Matts on 2002-03-16T16:59:34
Sorry, I didn't mean you underestimated the work involved - I actually meant the people doing the work do.
KDE is great - I'm really happy with it's consistency (I'm using 2.2.2 on my laptop) compared with my desktop OS (Ximian Gnome). I think sooner or later those issues remaining are going to be ironed out just fine. It just takes time, and even in commercial endeavours sometimes cool features take over from usability (remember Corel Draw's rollup windows as an example from the Doze world).
The hardest thing in the world though is to be an open source developer and take the "knocks" from the end users as they demand that your software be better. It's simply not possible to devote those kind of resources needed away from your personal life all the time.
(and yes, I promise I'll get AxKit 1.5.1 out REAL SOON NOW;-) Re:Hard work
ziggy on 2002-03-16T18:37:23
I think sooner or later those issues remaining are going to be ironed out just fine.Exactly. The only issue is how long "sooner or later" will actually take.
:-)
I read A History of Modern Computing a few years ago when it came out in hardcover. The author spoke at DC.pm about a week and a half ago. One of the long-term themes or cycles that seems apparent to me[*] in software is that the long term trend for software development is towards open source. That's not because open source is fundementally superior to all else on the planet. It's more a result of software being about solving problems, and eventually yesterday's hard problems become today's homework assignment for undergrads in Comp Sci.
How does this trend impact open source office apps? What we have today is reasonably comparable to the earlier office apps that were around in the late 80's and possibly early 90's (a gross oversimplification on my part here). What's missing is the knowledge and deep understanding of the problems that have been solved in the last 10 years or so being reflected in KDE/KOffice.
Sure, lots of those issues are in the process of being resolved, but they're not simple applications of garbage collection, compiler theory or interface definitions. They are hard problems to be solved, and they aren't the kinds of problems that big beefy Comp Sci curricula focus on.
But we will solve them. And we need more feedback from people like Cathy.
*: Of course, I'm perfectly willing to accept that I'm being delusional here.
:-)
