Home
Directory
Frontier
DaveNet
Mail
Search
Guestbook
System
Ads

News and commentary from the cross-platform scripting community.
cactus Mail Starting 6/24/97

From: tebbs@headspin.com (Brice Tebbs);
Sent at 6/24/97; 6:49:50 PM;
The trouble with flash

I agree with you that buying into a Java VM is no way to solve the problem that HTML sucks for multimedia. However, I am not sure Flash is the answer. Do we want to build our content in a format controlled by a single company? How about developing an open format for Flash like data. The all the content creation apps will create it. All we need is a simple format or scripting language with powerful graphics features. Oh yeah, and a few million dollars to hype it.

From: lineTo@gak.com (Greg Kimberly);
Sent at 6/24/97; 6:49:18 PM;
Re:LineTo MoveTo

Besides Starlogo, Mitch Resnick, Brian Silverman and Andy Begel have a version of Logo available that runs on Java. Folks wishing to take a trip down memory lane (or looking for a programming environment for their kids) should take a look at:

http://bongo.www.media.mit.edu/people/bongo/

Personally, I think the annoying thing about all the available browser programming tools is that they don't let you muck with the content in the page. I want to draw lines from (the bottom-right of the first image) to (the top-left of the third image), not inside some box. What makes web animation so sterile is that it's off in its own little corner...

From: M.J.Jennings@damtp.cam.ac.uk (Michael Jennings);
Sent at 6/24/97; 11:30:44 PM;
Re:LineTo MoveTo

> We were doing this in the 70s. In the 90s it isn't rocket science.

Someone has finally noticed. Thank goodness for that. I don't know if it's just that the industry is full of people who were barely born in the 1970s (if they were), but the tendency of the industry to reinvent the same wheel over and over again except requiring a 200MHz pentium and 20Mbytes of RAM is kind of depressing. And of course the press' ability to talk about it in hushed, isn't this amazing, tones is even worse.

From: tbyfield@panix.com (t byfield);
Sent at 6/24/97; 2:51:11 PM;
fwd: Edward Tufte on the World Wide Web

> From: glen mccready <glen@qnx.com>
> Forwarded-by: Keith Bostic <bostic@bostic.com>
> Forwarded-by: guy@netapp.com (Guy Harris)

Tufte's the guy who wrote *The Visual Display of Quantitative Information*, and seems to know at least a *little* bit about good visual design :-); in an interview in the Computer Literacy Bookshops New Book Bulletin, he said:

I can tell you something else about the poorly produced sites; in their designs, the allocation of space on the screen tends to reflect the distribution of the political power controlling the site. Programmers have a great deal of control, so there are lots of fancy tricks employed... designers control a great deal, so there are elaborate page navigation systems, and elaborate buttons to click on. The result is that content winds up with only tiny share of the screen, often only 20-30% of the bandwidth! The rest is computer or administrative debris, or over-produced, over-crafted buttons.

The full interview can be found at

http://www.clbooks.com/nbb/tufte.html

on Computer Literacy Bookshops' Web site (which is probably fairly well designed by Tufte's criteria -- and lets you browse their catalog database and order books over the Web).

From: rbrandi@lucent.com (Ralph Brandi);
Sent at 6/24/97; 5:22:43 PM;
moving text

Why do site authors persist in thinking that users want moving text? I *don't* *want* text that moves. I want text that stays on the page. If it moves, I have to chase after it. I don't want to chase after it. I want it to be in one spot. If I look away, I want it to be in the very same spot when I return.

Philip Greenspun just released an excellent book called "Database Backed Web Sites, The Thinking Person's Guide to Web Publishing". It's one of the most readable computer books I've ever read; I finished it over last weekend. You can read parts of it at:

http://www-swiss.ai.mit.edu/wtr/dead-trees/index.html,

where it goes by his original title, "How to be a Web Whore Just Like Me". One of the chapters in his book is called "Java and Shockwave -- the <BLINK> tag writ large". Unfortunately, this chapter isn't online yet, but if memory serves, he basically makes the point that I made above (except much more wittily). The <blink> tag was a joke. I thought it was even a little funny. But Java and Shockwave and the 14 million other programs that want to make your text nervous aren't jokes. (Well, okay, maybe the original "Nervous Text" Java applet is too....)

I hear from my clients all the time that they want a little motion on their pages, that they want scrolling, that they want the all singing, all dancing, all flashing Rockettes-inspired dazzling Show-To-End-All-Shows. Meanwhile, all the users want is to be able to find and digest some information or entertainment.

Books don't blink. They don't belch, they don't barf, they don't bleep. Their text doesn't move. Once you get past the age of about 4 or 5, they don't make sounds either. Maybe that's because we don't need the distraction.

Steve Falkenberg just released Web Free, which I believe you've even mentioned on your page. Yeah, it's nice that it blocks cookies and ads, but the best thing to me about the software is that it stops animated GIFs from looping. I can't stand to read an article about something I'm interested in and have my eye constantly drawn away from the article to somebody's stupid blinking, flashing, animated Cool Thing. I have to work twice as hard to get the information out of the article. Not good. Thank you, Steve.

Brandi's Law of Web Site Design, number 1: If you don't have a reason for including glitz and flash, don't. If it doesn't add anything to the users' ability to comprehend your message, it has no place on the site.

If Flash makes it easier to include gee-whiz effects on a page without causing authors to reflect on whether the effects really add anything to users' comprehension of the message, then it does the web and its users a disservice.

From: mikem@esd.sgi.com (Mike Mace);
Sent at 6/24/97; 12:32:27 PM;
Re:LineTo MoveTo

You hit a nerve for me -- I started out as a PostScript programmer, and have never been happy with the way graphics are developing on the Web. I think some sort of description language, rather than tags and bitmaps, is much superior way to present graphics in a bandwidth-constrained environment like the Web.

An analogy is early desktop publishing with PostScript printers. Why did those printers sell so well even though HP LaserJets at half the price could print bitmaps at the same resolution? Because it took forever to send a 300 DPI bitmapped page over the LAN, whereas you could often get better output with a 5k chunk of PostScript instructions sent in a fraction of a second.

I think the Web needs a simple, speedy, open PDL (which leaves out Acrobat, IMO). If Flash can become that, great. If not, one needs to be defined.

From: webworks@sirius.com (webworks);
Sent at 6/24/97; 11:59:50 AM;
Re:800 Words on Java

I understand the dismay surrounding Java. I guess we all expected a backlash.

But... Java is actually even more real. I was never a giant Java fan until recently. What changed my mind? Server-side Java programming. Go check out the JSDK from Sun (it's at http://java.sun.com/ of course). It allows you to write server side programs that run on an platform that has a Java Virtual Machine, with Netscape servers, Apache servers, IIS or (of course) Jeeves. It has utilities to abstract much of the common work server-side programmers are faced with (parsing GET and POST data, managing cookies, etc). Now check out the new JDBC stuff. Drivers for most every database you can run. Easy abstraction of SQL. Powerful. All of a sudden writing server-side programs for Web servers has become a LOT easier (and quicker). And, of course, your code is portable and reusable. Brilliant.

Also check out Dynamo http://www.atg.com/. Sick middleware for your complex site all in Java.

Applets shmapplets - marketing people just use them to create spinning balls. boring. the real power, the real magic is starting to happen on the backend.

From: tomalak@dsoe.com (Lawrence Lee);
Sent at 6/24/97; 11:54:40 AM;
Flash, XML and Special Stories Site

Flash: Near term solution... vector graphics is great for a lot of things and save on the bandwith.

XML: Maybe next year? It looks promising, but will the browser companies support it... Microsoft has been talking about it, what about Netscape?

From: roy@sundahl.com (Roy Sundahl);
Sent at 6/24/97; 11:51:38 AM;
Re:LineTo MoveTo

I've read your column for a while now. Good stuff. I cut my teeth on the KIM-1, the NorthStar Horizon, the UCSD P-System, PDP-11, Oregon Pascal, etc..

Back in 1987, I wrote a color Mac ToolBox for DOS and UNIX back by implementing most of Inside Macintosh Volume 1 and utilizing a product called MetaWindows for the QuickDraw part. MetaWindows waw written by one the PARC guys that DIDN'T wind up at Apple! I shipped two "Mac" products on DOS and UNIX before Apple had a color Mac out. It read resource files created by the Mac directly so you could use Mac tools to build your UI.

I did WindowManager, ControlManager, DialogManager, EventManager, MenuManager, etc., even ResourceManager and FontManager. Once FonManager was done, I could use Chicago-12 and let me tell you, it was pretty fun seeing that on a DOS machine! I'm telling you this because you can appreciate how fun it was to reverse engineer the whole toolbox. The point is, I was hooked on the Mac from there on because I really got a good feel of how cleanly things were working in there.

Anyway, in response to your article, I wanted to tell you about a little project that I did for Lari software a few months ago. They have a plug-in called Electrifier which basically sends QuickDraw-GX shapes (which are usually very very small) to a user and the shape is animated on the local machine. GX can do some real nice stuff. Too bad about its prognosis.

So, I did the Windows version of Electrifier since I had ported QuickDraw-GX to Windows for Apple previously. I think that QuickDraw-GX is the kind of API you'd like to have available in the "ToolBox" of today. Just thought you might be interested.

From: dwiner@well.com (Dave Winer);
Sent at ;
Re:LineTo MoveTo

With vector graphics we can both have beautiful lo-bandwidth sites.

That was the point I'm making. No can of worms. Just a feature that hasn't been implemented by the browser vendors.

Dave

From: cameron@michweb.net (Cameron Barrett);
Sent at 6/24/97; 2:37:52 PM;
Re:LineTo MoveTo

Yikes, It seems like Dave has opened a can of worms.

I am quite torn between the two opposing theories. As a designer, I want everything to look pretty and as a net advocate, I want everything to download fast. Dave's gripe is the text-GIFs we see on many sites, including my own. Personally, I do not like the default browser fonts (Netscape-Geneva, Explorer-Times), so I create small text-GIFs in Photoshop and use those as "buttons." Each one of these text-GIFs is less that one or two kilobytes. Some are as little as 300 bytes. This is a better solution (for me) than a table with background colors as we see on Dave's site.

I can easily see how people who are programmers (and have analytical-type minds) would be in favor of an all-text HTML page, very similar to Jakob Nielsen's AlertBox or Dave's own Scripting News. They download fast (even on a 14.4K modem) and the information is easily accessible, almost immediately. For a programmer, this is a dream come true. However, for a designer (who thinks more visually), this is not good. The sites begin to all look the same. There are only a few basic layouts to choose from, and creating a corporate identity or brand is nearly impossible.

So, what does it boil down to?

You have to design a site around your majority audience. For a site that receives a lot of visitors in the programming realm (like Scripting News), a text-only fast-loading page is perfectly fine. The programmers are not there to see pretty pictures and don't want to wait for them. However, for a consumer-based site, pretty pictures and graphics are almost a requirement. When you hit a consumer-based site, you are almost expecting to be entertained, much like when you watch commercials on TV. A boring commercial is just that. A boring (read: no pretty graphics) web site follows suit.

Vector-based development tools are a good thing. It gives the designers a way to make pretty graphics that download quickly.

Site references: http://www.useit.com/alertbox/ http://www.cambarrett.com

From: ccline@sbforums.com (craig cline);
Sent at 6/24/97; 11:15:00 AM;
Re:LineTo MoveTo

The most bizarre aspect of the "browsers wars" is that in their competition to make their browsers the be-all-end-all, browsers are growing in size, becoming increasing CPU hogs (I regularly get the message on my IE4 beta "A script is hogging the CPU - sh ould it be stopped?"), and are able to correctly display fewer and fewer sites. Yes, IE4 is beta, but geez, you'd expect to be able to access Mcirosft's web sites, but Nooooooo. More often than not I hit a site with it, it reads in the files, it says don e, and I'm left with a white screen. So sock it to them!

From: reede@willriley.com (Reede Stockton);
Sent at 6/24/97; 10:24:55 AM;
Re:LineTo MoveTo

Let's get real?

You're missing the point. The idea which is pushing Java isn't how to get there the most efficient way with today's system -- it's how do we grow the market in a way that frees us from the constraints of platform limitations.

Hardware is cheaper than software development. Always has been. Always will be. Bandwidth will continue to grow. Local computing horsepower will continue to grow. NCs are a realistic option. TVs with keyboards, Java chips and network storage are a realistic option.

Do we want to be able to deliver stuff today? Sure. Can we do it today with Java? Probably not. Does that mean Java is bull? Sure. And the 128K Mac was a toy.

Ignore Java at your own peril. Let's compare target market size in 5 years.

From: dwiner@well.com (Dave Winer);
Sent at ;
Re:LineTo MoveTo

Tim: No way! I have a full T1-connection and the fastest Pentium they make. It's way too slow for me. This piece is right, everyone is agreeing (see the mail page). Why not have XML operations for doing graphics right in the browser? I bet you could do it in 30K. No plug-ins, no Java. And cellphones and handhelds won't have high-speed connections for a long time. Dave

From: tlundeen@lundeen.com (tim lundeen);
Sent at 6/24/97; 10:10:57 AM;
Re:LineTo MoveTo

You seem to be assuming that network delivery speeds are going to stay at 28.8, but they won't. Within a very short period (we might argue whether this is 1, 2, or 3 years) the norm will be a true high-bandwidth connection to the world, especially for downloads. At that point, today's "slow" Java becomes instantaneous, and the cababilities of a full language become invaluable.

I agree that we are on a point of the curve right now where the user experience isn't ideal, but today's Java-based solutions are a good solid step on the path towards a high-bandwidth world.

Flash is a short-term solution, meeting some of today's needs. But Java also meets needs today, and will play a bigger and bigger role as network speeds increase. With a fast network and the right Java libraries, you won't need Flash anymore, but Java will still be with us.

From: chris@aggroup.com (Chris Daily);
Sent at 6/24/97; 10:04:37 AM;
Re:800 Words on Java

I'm begining to think Java's place in the Internet is a lot like C's place in the Unix world.

C was supposed to be Unix's cross-platform language. But it didn't end up there. Instead it was the least common denominator. Posix & Motif brought a noble attempt at cross-platform support for OS type services, but everybody wanted to provide something better, so they embraced and extended. The result is that writing a 100% Posix complient application is tedious and time consuming. There are always glitches and it's still the least common denominator. Somebody always comes out with something new.

Java is a lot like this. It's a great language and offers a large set of OS services. The assumption of the write-once-run-anywhere cult is that these services are enough. They're not. Even under the assumption that blatent problems like printing and speed will get fixed, it's still not enough. People want to embrace and extend. They need to. It's human nature.

Unix has become fragmented and complicated, but it's still a great platform. From a engineer's point of view it's extremely productive, stable and fast. I can write an application which takes advantage of special platform specific features, but still know that porting to another platform wont be that big a deal.

I look at Java that way too. My GUI's are easy to build, the language is super productive and any work I do in Java is more or less platform independent. I can now think about writing real applications for Mac, Windows and even Be boxes. That's great!

From: jonathan@pathfinder.com (Jonathan Hirschman);
Sent at 6/24/97; 12:43:09 PM;
Re:LineTo MoveTo

I think you're on the right track, but betting on the wrong technology. I also take some issue with your feelings about Java and lightweight graphics/animation/effects.

I'd suggest that you take a look at two technologies:

Macromedia is probably the most "closed" technology of the three, since its Java based player doesn't work terribly well (which is why they aren't pushing it). This technology is unlikely to work on NCs, appliances, palmtops/PDAs in the short-term future.

From: dave@sherm.com (Dave Sherman);
Sent at 6/24/97; 9:39:55 AM;
Re:LineTo MoveTo

Check Seymour Papert's book, The Connected Family. Papert, who worked with Minksy on AI and Piaget on childhood development, worked on Logo as a tool for childhood exploration. His book provides a cogent explanation of the dysfunctional attitudes that adults have towards computers. Worth a look.

Also check out Negroponte's column in this month's wired. I think your beef with web browsers extends to the whole PC experience. I think the technogeeks may be spending more time admiring their riffs that in playing something that is useful to the general populace.

Scientific American also has a good article this month on how the latest gee-whiz technologies aren't helping business productivity.

I wish someone like MS or Intel could effectively market simplicity and elegance. Lord knows Apple can't do it.

From: diana@acmetech.com (Diana Oswald);
Sent at 6/24/97; 12:33:27 PM;
You got that right!

Oh Dave, your last DaveNet really struck a chord with me. WWW has come to mean the world wide wait for me. I got so tired of it the other day that I even shut off my graphics. What do I want when I surf the web? I want information. If I wanted to see pretty pictures, I'd go to an art museum.

From: mslade@ix.netcom.com (Michael Slade);
Sent at 6/24/97; 9:24:42 AM;
Re:LineTo MoveTo

Of course, what you refer to as turtle graphics came from the Logo language at MIT. As you know, the more common term is vector graphics. For great vector graphics, look at PostScript. The print industry does.

When PostScript first became available in the first Apple LaserWriter and before there were great tools for it I coded PostScript by hand. It was like Forth (which I already knew) but gave me beautiful, resolution independent vector graphics. When tools became available (Illustrator, Freehand, all desktop publishing tools, PhotoShop...) of course I stopped hand writing PostScript code.

For the most modern version of this look to the upcoming 2D api for Java being codeveloped by Sun and Adobe. This will become part of the base standard of Java and viola - great graphics in a publicly documented standard.

I spent many hours this past weekend studying Flash 2. Flash's vector basis makes it far stronger than Director as a sign of where things should go. The way to get there is Java and the 2D api.

The problem with Flash 2 is that it is a proprietary Macromedia tool combined with a proprietary file format and nobody else will make tools for it.

Adobe creates great infrastructure standards (such as PostScript and Acrobat) and graphic tools while Macromedia makes great media integration tools. I'm looking forward to the day when Macromedia can make great tools for Adobe's infrastructure.

This solution is already in the making and should just work with both Microsoft and Netscape browsers because of their built in Java support.

I hold no stock in Adobe or Macromedia - I've just been using their stuff for a long, long time.

From: adul@cmg.FCNBD.COM (Albert Dul);
Sent at 6/24/97; 11:14:59 AM;
Re:LineTo MoveTo

When I saw these PostScript operators as the name of your column, I thought "Cool! Someone of influence is trying to get some cool imaging tech hooked up to the web!" I don't know if it could actually be PostScript. I guess Adobe would own the whole place . Still, it only takes a tiny bit of PostScript to produce nice images.

Here's a nice gray gradient, 400 pixels wide: 

	20 2 200 {
		dup 200 div setgray 0 exch 20 sub 400 2 rectfill
	} for

How long would it take to transmit this? Don't blink. You can tweak the above code and define it as a procedure. Send it once. Call it many times with different colors and dimensions. You've sent a tremondous amount of graphic information in a hundred byt es or so. Yes, this is a simple loop, but it is representative of what can be done.

There's certainly no shortage of PostScript illustration programs out there, but I don't know how small the files they generate are. I suppose one could create a library of predefined PostScript procs that allow a few parameters each to customize the look .

To use today's buzzwords, it's kind of like Java for graphics: platform independent as long as you have an interpreter that runs it.

There used to be a windowing system for Suns called NeWS, for Network Extensible Windowing System, written completely in PostScript in an object oriented Smalltalk like environment. It was very cool and very threaded (click on a menu and things still run! A menu ran in its own thread) and never caught on.

An app could run on your machine and the interface could appear on my machine by shuttling the display code back and forth. NeXTers know this as NXHosting.

From: andyjw@tiac.net (Andy J. Williams);
Sent at 6/24/97; 12:02:42 PM;
Re:LineTo MoveTo

Turtle Graphics were born from LOGO. LOGO was the brainchild of Seymour Papert who heads up the Epistemology and Learning Group at the MIT Media Lab. His view was that computers could help children learn math by having them become PART of math. Want to understand a square? Walk it out on the floor. Take four steps forward, turn to your right, take four steps and so forth.

Telling a turtle to take these steps was a way of doing this using a computer (though he originally started with crude robots back in the 60's).

Today, LOGO is still used in many schools and has varying rates of success.

One of the more interested innovations was to create programmable bricks for Legos. Now kids can use turtle commands to make Lego-based robots wander around a room. With some enhancements (light or sound sensors) kids can program very bizarre and interesting contraptions.

Mitch Resnick (in the same group at MIT) has designed a new LOGO called StarLogo which replaces the single turtle with thousands of turtles. Suddenly it's a programming tool which allows children (and adults!) to explore group behavior. It's very simple to program basic ant behavior and then let these ants loose on the screen for an ant colony simulation. Many other simulators are possible including traffic, gas molecules, and so forth.

In short, the turtle graphics idea is still used in education with interesting results.

See the directory site for a list of important pages on this server This page was last built on Wed, Jun 25, 1997 at 7:27:21 PM, with Frontier. Internet service provided by Conxion. Mail to: webmaster@content.scripting.com. © copyright 1997 UserLand Software.