Joel and Norman Totally Miss the Point on Simplicity
Norman wrote a overgeneralized post about the vices of simplicity to which Joel responded. Both of them have the premise that the most important part of simplicity is features. Joel even goes on about how everyone comments how simple the iPod design is, yet Joel tells us that is just aesthetics, as if aesthetics doesn't sell.
Simplicity is not how many features your product has, simplicity is an idea that can be applied to all parts of a product.
The user interface, the value proposition, the target audience, the infrastructure, the way its components interact, the default preferences, the way in which you change preferences, the way in which you communicate upgrades, the way the product interacts with other products, the api, the manner in which you pay for the product. This list can keep going, these are only a few examples.
Joel and Norman claim that complexity is better to strive for than simplicity, but I think that they are making too broad of a statement. Does anybody really think that the user interface should be more complex than needed? That the target audience be an amorphous void? Their main point seems to be that more features are always better than fewer, however, even that is an over-generalization.
In a startup environment for example, if you strive for more features instead of fewer, that could mean your startup fails to launch and without a product you have nothing. Beware of this over-general advice, these people like to write inflamatory un-predicated articles with no thought to the repercussions.

11 Comments:
I also think that perhaps their view is company centric rather than user centric. From a company's POV, you make more money if you add more features, but from a user's perspective, you just get really pissed off and burn the company's headquarters down. Or something. Either way, their view is correct if you don't care about your users/audience.
2:53 PM, December 12, 2006
I think you need to make a distinction between elegance and a "feature fast".
Elegance is good. Telling customers they can't have the features they want because you know better than then (which I've heard over at 37s) is bad.
It is true customers only need 20 of the 100 features available. The solution is to develop a system allowing the user to PICK the 20 features THEY WANT from the 100, using a software product line approach to give everyone their own generated app.
That is the way we are going over at SystemsForge, and I think we're in the vanguard of what could become an important trend.
Best Wishes,
Peter
8:44 PM, December 12, 2006
I think that you missed the point. They both were, very specifically, talking about simplicity or complexity of the user interface presented the user, and the features it offered. They claim that presenting a more complex user interface will probably win over a simpler one. And Joel already addresses the "startup environment" that you talk about.
2:32 AM, December 13, 2006
If they were so specific about talking about user interface, why is it that the words "user interface" do not show up in either post? And the heart of my point is that both of their posts are over-general. Your summation is over-general: "presenting a more complex user interface will probably win over a simpler one". More complex or more simple in what ways? Win in what context? Are radioshack products a good way to measure how people react to computer based products? What are the pros and cons of this comparison or does it not hold at all?
Both articles talk about a complex user interface... or one with more features... as the only important aspect of simplicity vs. complexity and I am pointing out that that comparison is only one of many and that what sounds like it might be a deep meaningful piece by those two turns out to be a thoughtless half-assed rant.
5:40 AM, December 13, 2006
What gets me is that Joel backtracks by saying the iPod is not simple because it has asthetics as a feature. Basecamp is not simple because it has "well thought out" as a feature.
When actually both are simple because they do what they need to do, no more, no less, with an option to extend if you need it (and essentially are willing to pay extra to have that simplicity mucked up)
5:24 AM, December 14, 2006
I see 'more features' as an inevitability in any products life. All things being equal (an always dangerous proposition) users would choose to have a feature rather than not. People like the thought of more functionality.
As people mature with a product, they become adept at using it. This usually means that a product, especially the interface, will grow with them. The problem becomes, when your mature product has a powerful but intricate interface that seasoned users love but new users find unwieldy.
Few examples of this:
- game console controllers
- mobile phones
- remote controls
- Yahoo's home page (remember how simple the early versions were)
So to me, the difficult question is: how do you satisfy your current userbase without alienating new customers?
8:38 PM, December 15, 2006
Seems like they both over generalize. Joel conveniently ignores that the iPod doesn't carry an FM tuner, for example.
In any case, simplicity is, IMHO, not a matter of less features as much as the way features are accessed and implemented. I think most people will agree that OS X is simpler to use than Windows XP. Is it for lack of features?
5:52 PM, December 18, 2006
“Beauty is more important in computing than anywhere else in technology because software is so complicated. Beauty is the ultimate defence against complexity.”
-- David Gelernter
5:11 PM, December 19, 2006
With your Ruby starfish parallel framework, can I use it to distribute the subprocess task to another machine?
For example using mongrel connect to a cluster of computers to process multiple subprocess parallel tasks?
1:44 PM, January 16, 2007
Of course.
1:47 PM, January 16, 2007
>> Starfish used over multiple machines.
If you don't mind explaining, how might I do that?
9:43 PM, January 16, 2007
Post a Comment
<< Home