Order Toll Free 1-877-339-5855
News, How-tos, and assorted Views on Accordance Bible Software.

Tuesday, April 25, 2006  

Easy is Hard

In yesterday's post, I talked about the Accordance Construct window, a graphical interface for constructing sophisticated searches which has been around since version 1.0. In addition to explaining how easy it is to understand, I went on to contrast the Construct window with other approaches to building searches graphically. And naturally, I argued that the Accordance approach is superior.

In response, Mike in Siberia(!) left the following comment:

You're right — Accordance has something going there. But why invite flak! Why not just extol and demonstrate Accordance's powerful graphical search abilities, instead of setting a tone of 'my graphical search engine is better than yours!'

Besides, as you said, "It is not rocket science."

I can certainly see where Mike (and others) might have thought I was being too competitive, so let me clarify my intentions here.

First, I had no intention of denigrating the efforts of other developers. My goal was simply to point out that there is a fundamental difference between the Accordance Construct window and all other "graphical search engines." While I gave a brief synopsis of the history of graphical search engines, I did not name any other developers; nor did I include any links or screenshots of other graphical search engines. I even wrote that the free-form system developed on the Windows side does make possible "a few things which the Construct window [doesn't] do." I actually applaud the first developer I mentioned for developing their own system rather than just copying ours.

Second, my tone should not be read as saying, "my graphical search engine is better than yours!" for the simple reason that I am not addressing those other developers directly. Rather, I am writing to an audience made up primarily of Accordance users. They, rather than our competitors, are the "you" I am addressing. So my tone is better read as, "The graphical search engine you [our users] have at your fingertips is clearly the best, and here's how. . ."

Why can't I just tell our users that our approach is "the best" without contrasting it with other approaches? Partly because everybody claims to be the "best," so users rarely pay attention to such claims. But more than that, I want to give our users a basis for comparison, so that they will be better equipped to evangelize others on the merits of Accordance. If you, our users, start extolling the virtues of the Construct window to a colleague or friend, and are answered with, "Oh yeah, brand Y does that too," I want you to be able to point out that not all graphical search engines are created equal.

I'm not quite sure what Mike meant when he wrote, "Besides, as you said, 'It's not rocket science.'" When I wrote that "it's not rocket science," I was talking about the Accordance Construct window, not about graphical searching in general. What's more, I meant it from the perspective of the user, rather than from the perspective of the developer.

When I first got a Mac way back in 1992, I read a book called The Macintosh Bible. In that book, there was a chapter entitled, "Easy is Hard." The point of that chapter was that the Mac is easy to use because the developers have done the hard work of making it easy for the user. It's easy to make an OS or software program that is hard for the user to use; but very difficult to create something which the user finds easy.

Thus, from the developer's perspective, creating a graphical search engine is rocket science. You have to figure out how to represent word relationships using pictures, how to make it easy to understand for the user who may not happen to be a mathematician or logician, and how to prevent the user from creating combinations which are logically impossible. Every developer who has undertaken such a task should be applauded; but that doesn't mean that every approach is equally successful. I would argue that Accordance has been the most successful at creating a graphical search engine which is not "rocket science" to use. In the posts which follow, I'll attempt to prove it. :-)

Since you take the high road in this follow-up (and I really appreciate that!), I'll observe that Macintosh is not just a platform or an operating system, it's also a discipline. The Macintosh interface started with rigidly-enforced "human interface guidelines," which ensure that users who learn how to operate within one application (even the Finder) will have learned the great majority of what they need to know for operating all applications. I can remember the first Mac I unpacked (a Plus), and how it offered a tutorial in how to use the mouse, what to expect in the menus, and how every menu in every application shared common elements (such as Edit, with Undo/Cut/Copy/Paste/Clear always appearing at the top).

The fruit of these guidelines is that a developer who chooses not to follow them has historically been hooted off the Macintosh platform (or had to do major rewrites of their software). Early Windows applications reveled in being individualistic -- a holdover from DOS days when all interfaces had to be invented from scratch. But, a developer grounded in the human interface guidelines approaches their application the way David says -- knowing that "easy is hard" but choosing that hard road for the sake of the user.

For which, we thank you! :-)

Okay, I concedeā€¦ I was too hard in my comment yesterday, and I apologize. Thank you for your follow up. Your points in this post are valid.

Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?