I was doing some file system housekeeping and came across this little rant (below - with a few typos corrected) I wrote just over 4 years ago now and posted to Russ Nelson's Free Software Business (firstname.lastname@example.org) mailing list. I was surprised how quickly it was picked up by people such as Linux Weekly News (LWN). Apparently it struck a cord among some people. It also led to an interesting discussion on linux-elitists.
From: Larry Augustin <email@example.com>
Subject: Free software as a replacement for Microsoft
Date: Sat, 15 Sep 2001 14:02:19 -0700
I've been lurking on this list for some time, and some of the things I hear seriously bother me. They bother me because nothing I've seen on this list indicates to me that people here have any real concept of what users like about Microsoft products or what they do with them.
How many MS Excel, PowerPoint, Word, Outlook, or Access gurus are there on this list? If the goal is to displace Microsoft, we must be experts in Microsoft technology. We must understand Microsoft inside andout. We must understand how and why people use it.
We must put ourselves in the mind-set of the *user*, not the mind-set of the *developer*. Remember, the user doesn't know anything about code. The user doesn't write software. The user just wants to get their job done.
The free software community was able to recreate Unix because it understood Unix from its applications to its kernel. The free software community was Unix users. They understood how and why people used Unix. I see no one in the free software community that similarly understands Microsoft software. If we don't understand it, how can we displace it?
I recall a discussion (not on this list) some time ago where a group of people were arguing that Gnumeric was a replacement for Excel. I was appalled. They were arguing about Excel vs. Gnumeric features. They were arguing about reading and writing Excel file formats. They
didn't understand why Excel users complained when they tried to use Gnumeric. The prevailing opinions were that users were just not willing to learn to use something different.
I finally asked the question, "Can Gnumeric do pivot tables?" I got the response, "What's a pivot table?" My point was proven. The Gnumeric advocates didn't even understand the technology they were trying to replace. I can hand an Office power user an Excel spreadsheet with thousands of names and addresses, and with a few point and click operations, out come pages of stick-on mailing labels. How do you do that with Gnumeric? I'm willing to bet that few or none
of us on this mailing list have that level of proficiency with MS Office or Excel. If we don't know what it can do, and we don't know what people do with it, how can we replace it?
I recently switched my wife from Netscape on Linux to IE on Windows. Why? Because Netscape is *awful*. It's slow, crashes often, and too many things just don't work. Any impartial analysis would tell you that IE is vastly superior to any other browser from a users point of view.
I'm not sure that free software people get that. To free software people, the *how* seems as important as the *what*.
In order for free software to succeed, it has to be *better* software, where better is defined by the user, not by the developer. Again, this is about the what, not the how. Netscape could have an elegant architecture underneath. IE could be crap. IT DOESN"T MATTER.
Again, we must put ourselves in the mind-set of the user, not the mind-set of the developer. Remember, the user doesn't know anything about code. The user doesn't write software. The user just wants to get their job done. And their job has nothing to do with software.
You want to displace Microsoft on the desktop? Go spend some time in the shoes of a MS Office power user. Spend some time watching a true Excel wizard at work. Refuse to write a line of code. Learn what MS Office users really do. Until we have that kind of knowledge, we