News and commentary from the cross-platform scripting community.
	Broken Inside

By Wesley Felter, wesf@mail.utexas.edu.

In http://proceedings.www6conf.org/HyperNews/get/PAPER131.html, it is pointed out that Pointcast accounts for a lot of Web traffic. Various people have also pointed out recently that Pointcast's implementation fundamentally cannot scale to accommodate a majority of users on the Internet. The simple solution that all intelligent beings have known all along is multicast. Today it's not practical to use multicast because no one can get it (by the end of this article you'll know why). And no one has said that Pointcast won't support multicast at some point, but they have no incentive to do so.

Also look at Netscape Navigator, the Internet's favorite Web browser. If you know much about the Web, you know that Navigator is terribly deficient in some areas, particularly various HTTP options that can speed Web browsing by 35%. But Netscape has limited resources, so we can't expect them to support everything. It's apparent to me that they have a choice between supporting something like deflate compression and frames, or HTTP/1.1 and layers. If Netscape had added compression instead of frames, pages would look exactly the same to users, but would load faster. If Netscape had added HTTP/1.1 instead of layers, pages would look exactly the same, but they would load faster and use less bandwidth. See a trend here? The problem starts in the marketing department.

If Netscape advertised Navigator 3.0 as being twice as fast as 2.0 but with roughly the same features, why would people buy upgrades? As far as the user would be concerned, it's the same browser; you can't see compression or HTTP/1.1. Since Netscape is apparently run by the marketing department, they decided to skip technical improvements and add useless eye-candy features.

Do you use Mac OS? If so, you've probably heard of Mac OS 8, due this summer. Mac OS 8 is so great, according to Apple, that it deserves to be called a major upgrade. While Mac OS 8 does include a new Finder and some bug fixes, I don't think it's a major upgrade. The Finder will be faster, but until Apple fixes the low-level filesystem other applications won't get any benefit. The new user interface is nice, but I don't think it increases productivity noticeably.

It's the incentive, stupid

It's a sad but true fact that corporations today are run by marketing departments; what the company produces is what will generate the most revenue, despite what users think. Marketing being what it is, it usually does an excellent job of convincing users that they're getting something useful.

The naive solution would be to educate users. But who is to do this, the press? Again, consider Mac OS. If you ask a Mac user what he wants from future versions of Mac OS, he'll probably say preemptive multitasking, because the press has told him (repeatedly) that that will solve all his problems. But preemptive multitasking does nothing to keep misbehaving applications from crashing your Mac, which is the most common complaint. Protected memory prevents crashes, but the press too often gets the two confused.

So the next time your computer crashes or takes too long to do something, don't worry: whoever wrote that software is probably thinking the same thing, his hands thoroughly tied by the morally-impaired marketing department.