There is a very interesting discussion going on out there - out there where you all are - which is about a neat little tweek that you can give your Firefox browser. It has three levels to my mind but that is only because that is the full number of levels I can perceive at any one time. The tweek is explained here and was recommended to me via the Rukster. It has since been noted at Boing. I applied the tweek both at work and home and am quite satisfied that the alleged speed improvements have been achieved. I am happy with the change and that is the first level. But that is not the end of it.
First, there are messages being sent that by reconfiguring the browser you are not using it as it was intended which is characterized by the statement "remember that most of the defaults are defaults for a reason" from this blog. To be very fair, a few sentences up there is a clear invitation to experiement but it is noteworthy to me that there is also a bit of the old "we have thought a lot about this and the answer is no" that pervades Winerianly. I think this instance is merely a caution rather than in its true imperial form but there is a risk whenever the pre-thought out mantra appears that there is an end to improvement. That is the second level - doubt your own experimentation as the professionals have already been through.
The third level for me is the most interesting and is contained in the content of the same warning: "...[i]f you've enabled this, and you find pages that aren't displaying correctly, please don't blame Firefox or the Web developer." Steven describes the default setting's purpose here as:
It means a tiny delay before the rendering actually starts, but once it starts it will be smoother and more stable. Leaving this setting at the default can actually lead to a smoother-feeling browsing experience.It is my experience that people do not actually like smooth technologies over fast ones. Bank machines make a grinding pinging sound when your money is being provided to you for a reason. Something is happening you are being told. Similarly, I understand that gas pumps in Canada are intentionally slowed from their optimum speed because if filling your car does not take some small time, the customer does not get a feeling of value for money.
Most web developers work in small teams, including certainly those involved in open source development, and do not generally employ market research or sociologists. They experiment on those around them and make assumptions on their own experience - as we all do. But because they are performance geeks, there is a hanging presumption that that is normal. I think it is not or rather it is only one instance of normal. Sometimes. we like to hear our engine race a tad when we are sitting in traffic and we like to use elbow grease when cleaning the tub or to lean on the shovel when the tool is doing the work quite well thank you. We also sometimes like to see the web page reveal itself as a bit of work. We like to think something is happening and we like to think what we are seeing happening is leading edge, tinkery, a popular science experiment. So if there is a delay or a chunkiness that we can fiddle with, that is usually quite pleasant. Feels like the fine tuning dial placing us each in charge if only a little bit. That makes us participants rather than consumers.

Comments
Arthur - December 30, 2004 12:23 pm
So if there is a delay or a chunkiness that we can fiddle with, that is usually quite pleasant.
It reminds me of the 'progress dialog' dilemma: while that technique slows an application (abstract: 1. create dialog/window 2. show dialog/window 3. synchronize progress for visual feedback 4. free and destroy dialog ), people tend to think it's still faster than showing the 'hourglass/busy' icon alone.
Andrew - December 30, 2004 1:43 pm
As a web developer I can tell you that speed is definately preferred over smoothness in 99% of the customers I've dealt with. They want results, and they want them NOW - a 5s delay is usually perceived as far too long, no matter how complicated the background operation they just requested was.
Alan - December 30, 2004 1:53 pm
Once again, I am comforted by the pre-Victorian pace of my chosen profession: "we may be able to get this resolved before 2006" is not farcical.