bitsifter
friday, august 15 
[sift this]   The Java DeathWatch continued this week with the arrival of Frag Island, a Java based first person shooter style game.  It’s simple to play, just point Netscape Communicator to the page and, after a few minutes of downloading images and building maps, you’ll have a 100% Java based game with some stunningly rendered graphics.  Granted, it’s an early version of the game.  There are no creatures running around to shoot or other players to gib… actually, there is no real objective in the game at all.   Mostly, it appears to exist merely to prove that it can exist.

The Company was a buzz about Frag Island today.  Being keen on the idea that Java will succeed fuels the fire for marketing types when "proof in the pudding" Java applications:

See what Java can do?  Yes, I do.

Isn’t it amazing?  Not really.  Consider this.  Frag Island is based on Quake. -- a game which runs full screen, blazingly fast, and with 15 other players on a network on the Bitsifter 200Mhz Pentium Pro.  Frag Island is the end result of some serious engineering by some hard working developers, but Java simply isn’t up to the task.

But it’s cross platform.   Sure, Frag Island will run on any browser that has a Java virtual machine, but let’s dig a little deeper in the Frag Island site – take a look at the Platform page which describes how the application runs on a variety of hardware.

These results aren’t surprising.   The Windows monopoly forces Netscape to design the fastest and most reliable Java implementations for PCs, but what about Java’s cross platform promise?  If one platform is head and shoulders above the rest… if one platform represents 85% of the market…  why wouldn’t developers just write their programs for Windows and reap the performance benefits?

Fortunately for Netscape, large corporations have been buying into a lot of the Java promises.  These enterprises are full of a wide array of hardware and the idea that one language can provide applications for most of the desktops has them investing millions.  Better yet, they’re not that picky about performance as long as the solutions works.

These companies could learn a lesson from history.  Back in the early 80’s the University of California at San Diego had a language called the p-System.  The idea behind the language should sound familiar.  A developer would write a program for a “pseudo-machine” that existing only in memory and ran identically on different platforms.

Problem with the p-System and the reason why the early developers like Context MBA never saw their spreadsheet sell was the p-System was slow.   Compared to what?  Compared to Mitch Kapor’s lightning fast 1-2-3 spreadsheet written entirely in assembly language.

Today we have vastly more megahertz at our disposal than the PC XT that ran 1-2-3.  It’s easy to think these new processors will make Java shine, but I see no compelling reasons that native code will ever be slower.

[rant]   Access to the bleeding edge is costly.  It involves educating yourself in ideas that may make no sense in a matter of months.  Still, individuals are driven to the brink of technology simply because they want to know that they’re ahead of the game.   Problem is, there is whole lot of us.

Case in point: I’ve spent much of the day dealing with a "new" language called Perl.  Now, I‘m well aware of the fact Perl is up to version 5.0 and much of the geek community is aware of it’s existence, but indulge me, I had an epiphany and I want to share.

So, back to Perl.  Today’s exercise involved writing simple programs and I was stuck.  It was one of those “find any Perl programmer in the building and he’ll answer this in 5 seconds” type questions that drive me nuts.   Simple configuration issue.  So simple that a Perl reference manual wouldn’t lower itself to solving my problem.   Pure frustration.

It’s at this point that anyone trying to solve a problem thinks, “Dammit, how am I going to solve this?  No one is around to help since it’s Sunday.  Looks like I’m screwed.”  This is your ego talking, it’s telling you that you are unique in the fact that you’ve faced this problem.  It’s frustrating, but it makes you feel important.

Fact is, I guarantee that no matter how sophisticated your problem may be someone else in the world has faced it.  Even better, they’ve probably asked someone about.  At best, if the problem is technological in nature, they’ve asked their question over the Internet and, I bet, someone recorded it.

I had to believe in this fact when I went to DejaNews to figure out if anyone else had experienced my difficulty.   After a few brief queries to the database, I found roughly 50 response to the exact problem I was experiencing.  With that solved, I dove into my Perl pool of programming and continued my self-education happily.

Everyone has heard the collective knowledge in the Web is virtually unlimited.  I’m fond of telling of telling folks to “look it up on the Web” no matter what the question may be.  Still, the impact of this knowledge base has yet to infiltrate our lives in a way that we take it for granted that our shared experiences can make our lives easier.