News and commentary from the cross-platform scripting community.
|
||
|
Mail Starting 7/22/97 | |
Interesting comments on APPLE and flaming. I have often criticized Gil Amelio over the last five years. But everyone interpreted my negative comments about him as the result of personal animosity. They were right to the extent that I neither liked nor respected him. I would argue however that I reached those conclusions by trying to work with him for three years plus as Chairman of National Semiconductor.
From: PSprague@wavesys.com (Peter Sprague);
Sent at 7/23/97; 5:53:09 PM;
Re:Carl and the Two GuysMore public comments from National people would have been helpful in predicting Gil's tenure at APPLE. Recent comments from National Board members Charly Sporck and Gary Arnold reflect their true thinking. If APPLE had taken the trouble to call us before they hired Gil, they might not have hired him. I had lunch with Mike Markula after Gil was hired. I have known Mike for 20 plus years. He clearly didn't want to hear criticism of Gil so we discussed other things.
Flaming is not the same thing as criticism.
Hey dave, the web isn't HTML, it's URLs. And that isn't a new thought, btw, its Tim Berner-Lee's. A shockwave browser ... neat idea ... as long as it supports URLs, its part of the web. Also interesting notion that other formats like AOL's proprietary one, could also be "part of the web" if it was URLized.
From: sunilpaul@rocketmail.com (sunil paul);
Sent at 7/22/97; 5:15:31 PM;
Re:Challenging HTML...was the catalyst for the creation of the Free Software Foundation and the GNU tool suite? When Gosling was a student at CMU, he wrote a clone of Emacs (then available only for PDP-10 and Multics computers) for UNIX and sold distribution rights to Unipress Software. When Richard Stallman found out that there was a version of 'his' Emacs program that he could only use if he paid for it, it pissed him off to the extent that he started ranting about people hoarding software and impeding the free flow of code. From these rants sprung the first copyleft and the FSF.
From: marc@ranlog.com (Marc Ramsey);
Sent at 7/23/97; 1:27:25 PM;
Did you know that James Gosling...In college, my friends and I played a game like this. When one of us would start on a paper, the others would challenge him to include some word in it. It couldn't just be stuck in; it had to be used in a substantive sentence. I trumped most of the field by including "Velveeta" in a film paper.
From: chaz@visi.com (Chaz Larson);
Sent at 7/23/97; 3:09:06 PM;
Re:Trellix"Buttcrack" would have been a challenge.
I found this amusing, since I read it the day after I got my first pager, adding to the Pilot I already carry around with me, and the Swiss army knife on my keychain...now if I got a cell-phone too, I could get my first utility belt....though my co-workers would undoubtedly look at me strangely when I started jumping around the office yelling "I'm Batman!"...
From: phillip@webgenesis.com (Phillip Karlsson);
Sent at 7/23/97; 11:09:09 AM;
Information devices...I think that the other thing we're lacking in terms of ubiquitous information devices is more coherent ways to work with this information. I bought my Pilot because I got sick of trying to remember post-its to bring home with me from work with people's numbers, e-mail addresses etc. I'm continuously annoyed by the need to keep copies of people's information in Eudora at home, my Pilot synchronization database at home, and in Eudora at work...and anywhere else I might use it from. I'm waiting for someone to come up with an effective centralized storage mechanism for all this, that keep it automatically in one place. Put an API on that, and then let all the other apps just check that database, and focus on their own front-ends and extensions to that information. The main reason I don't want to start using Now Contact or some other contact manager, is that I have no desire to store all this information in yet another place...I'd rather deal with the shoddier interfaces on all the other products.
I recall the BeBox having something like this built into the OS...that's cool.
Re XML parsers in Frontier, yes, we've started. See:
From: dwiner@well.com (Dave Winer);
Sent at Wed, Jul 23, 1997 at 6:34:44 AM;
Re:"Trellix, HTML and Markup Languages"http://www.scripting.com/fatpages/97/05/cdfParser.html
Dave
Trellix looks good. If you've read any work by the designer Clement Mok, such as "Designing Business" (Adobe Press), you'll see the kind of information structure diagrams that look similar to the Trellix map.
From: phowson@t130.aone.net.au (Paul Howson);
Sent at 7/23/97; 1:10:10 AM;
Trellix, HTML and Markup LanguagesNetobjects Fusion (of which Mok is one of the designers) also has an innovative "structural" user interface which is clearer to understand than the traditional Windows or Mac interfaces. There is much room for innovation in this area.
With regard to alternatives to HTML, it's quite right that HTML is a mess, in large part due to confusion over its purpose. It was meant to be a structural markup language, but has been adulterated into a format markup language. Now it does neither very well.
We do need a good structural markup language --- even in the age of things like Trellix. As a graphic designer and publisher I know that providing the same information in either print or on-line form is going to be required for some time. How to do this translation with some degree of "automation"?
This is where structural markup is essential --- to identify the purpose of information rather than how it looks. SGML is the right idea, but it too has suffered from complexity disease. The newly proposed XML seems to have the right balance. Simple and sufficient yet powerful also. Isn't that the hallmark of good design?
BTW I think XML would be great in concert with a tool like Frontier for "compiling" web sites or indeed printed publications also. Is anyone working on XML parsers or translaters in Frontier?
Being a long time reader I think you'll dig this new piece of software from one of my old companies. It's called iVisit. It's written by Tim Dorcey who wrote the original CU-SeeMe and it lets you do point to point video conferencing WITHOUT CU-SeeMe's reflector limitation. I'm not a big fan of the company but I do have to admit this is going to open up the desktop video conferencing market to everyone since it runs GREAT on a 28.8 connection. Give it a look see.
From: jason@spewww.com (J. DeFillippo);
Sent at 7/22/97; 7:13:33 PM;
Pretty good new piece of softwareI still like you.
From: andyi@world.std.com (Andy Ihnatko);
Sent at 7/22/97; 12:13:43 PM;
Re:AttacksMake yourself a pan of Rice Krispies Squares and spend the rest of the afternoon alternatively nibbling on them and doing the "I Don't Care" dance.
I just saw Bricklin's mapping tool yesterday at Spotlight. It is by far the best way to get a view of a web site that I have ever seen. It makes more sense than Larry Tessler's demo last year of what amounted to lilypad-jumping. Bricklin has thought this out and has integrated it into such everyday metaphors as calendars and scheduling. It totally makes sense.
From: PatMeierPR@aol.com;
Sent at 7/22/97; 9:43:19 PM;
Re:Challenging HTMLYou speak of Apple as a myth, a ghost of the past that should be exorcised from our collective consciousness. Forget the Steves, let Apple fade away, explode the fantasy, get on with the necessary work.
From: dave@sherm.com (Dave Sherman);
Sent at 7/22/97; 3:24:31 PM;
Re:Carl and the Two GuysBut the Apple that appeals to me (and I believe others) is not the Apple of myth, but the Apple of reality. The Macintosh is the only platform that currently provides a capable, cost-effective alternative to Microsoft operating systems. Macs work well enough to stay at the top of heap in customer satisfaction and product loyalty. I don't care how Apple started, I care what they do now. And though the rest of the world seems to disagree, I think Apple makes a great OS and great machines. Tools that work great. Now. Today.
As a software developer, you offer a valuable perspective on Apple's troubles. But you and others from the technology industry may not appreciate what non-technical professionals face in the business world. The troubles lie not in the garage myth, but in the struggle for control between MIS departments and other folks who just need to get their work done. The Dilbert Reality has contributed more to Apple's woes than the legend of the company's origins.
To attribute the defense of Apple and the Mac platform to out-dated mythology does a disservice to those who find the Mac a useful and capable tool and who are perplexed that the rest of the world is armed with shovels around the anticipated-to-be-filled-shortly casket. Can you think of any other company that has great product, continuing innovation, loyal customers, billions in sales, and has pundits chanting last rites? I can't.
I think we expect too much from Apple. It is a flawed entity that reflects the best and worst of human nature, just as each of us do. Apple has made a significant and laudable contribution to computing. That's a fact, not a myth. I have faith that Apple will continue these contributions in the future. That's a dream, not a delusion.
GNU emacs. By far _the_ most powerful, and popular *NIX programmer's editor/lisp interpreter available at any price. I've been using it since before HTML existed.
From: toxic@wired.com (Jeff Burchell);
Sent at 7/22/97; 3:03:29 PM;
Re:Challenging HTMLBeing GNU software, it's free and comes with full source code. The source compiles cleanly under every Unix variant that I've worked on (I'm a sysadmin by trade, and have worked on most of them), and there is a Win95 port available (although most of our people use an X Windows emulator on their PCs or Macs, and then run emacs off a UNIX server). I believe there is a Macintosh port, but I don't know anything about it.
Details on other ports are available from ftp://mirrors.aol.com/pub/rtfm/usenet-by-hierarchy/comp/emacs/ (or your favorite rtfm.mit.edu mirror)
Emacs was designed as a programmer's editor that can be customized to behave appropriately with whatever language you happen to be working in. The HTML extensions that I use mark content that is contained in container tags
... through either indentation, or a color change. Makes stacking tables a piece of cake, and spotting missing closing tags (and the like) easy.But, this should be simple for an editor designed to automatically format code and notice a misplaced } or ; right?
Emacs is notoriously difficult to learn, but I've never met anyone who didn't have it down cold in a month (especially if they had access to an O'Reilly book), and I've never met anyone who switched back to whatever it was they were using before.
> I use simpletext exclusively (on a Macintosh) because it launches the quickest.As good a reason as any, as far as I'm concerned. I actually use vi for about 85% of my text editing work, for exactly the same reason.
100% of Wired/HotWired content goes through emacs at some point in the process. Our final production department refuses to use anything else, as do a good majority of the rest of our producers/designers.
Unfortunately, I can't contest that most Mac-based sites are done with BBEdit. But on the PC side, a lot of folks have used HotDog or, more recently, the very great HomeSite program. I bought Virtual PC so I can use HomeSite on my Mac at home (I've used it here at work for a little less than a year).
From: bobsoron@world.std.com (Bob Soron);
Sent at 7/22/97; 5:51:29 PM;
WYSIWYG WebI know what you mean here, but in my Web designing work, I've found it terrifically useful to be able to see what my tags are doing. I know BBEdit colors tags (so does HomeSite) but I like knowing that I've gotten the tags *right* without worrying about syntax checking. PageSpinner is about as close to this as I've seen on the Mac; I actually think it's a great HTML editor, but the file management tools are very lacking (and one of HomeSite's strong points).
I find it very ironic and sad that most savvy Mac users are so eager to throw out the platform's strongest points. Fron Day One, the Mac has had fonts, styled text, strong formatting. If you remember those early releases of MacWrite, they weren't always WYSIWYG either! Sometimes you'd look at the printout and screen and scratch your head. And now, the hottest Web editor on the Mac is a character-based editor that can display rudimentary formatting but is so *far* from WYSIWYG that it may as well be running on an 80x25 screen.
I know we all need generic text editors (I use Tex-Edit) and BBEdit's a decent program. But I'm baffled by the extent to which people are going out of their way to avoid the Mac's biggest strength! On the Web, every browser displays your page differently, it's true. But it strikes me as a very large non-sequiteur to decide that it follows that HTML editors with any formatting at all are bad.
As for PageMill, HomePage, and the rest -- I don't like them for creating pages. They create bad HTML and the pages they create are displayed even more erratically in browsers than hand-coded HTML. But they *are* useful for rapid prototyping of software that uses a Web-based front end. The company I work for is using the Web for transaction processing. HomePage lets me create three or four options for the interface for a page in a few hours. Once folks have decided what they like, I go in and hand-tool the code in HomeSite.
emacs? I use simpletext exclusively (on a Macintosh). I've used a number of WYSIWYG editors and I always go back to SimpleText. Why? Because it launches the quickest.
From: crawford@livepicture.com (Michael Crawford);
Sent at 7/22/97; 2:30:33 PM;
Re:Challenging HTMLAs you pointed out, HTML is tired and unwizzy. That's the way it was meant to be!
From: ivan@sven.ctnet.com (Drew Ivan);
Sent at 7/22/97; 3:37:33 PM;
Re:Challenging HTMLHTML was designed to be the best of both worlds. By using *logical* rather than *physical* markup tags, it would be a format that both humans and computers could understand. That would mean sacraficing wizziness, but it would enable computers to analyze, summarize, and index our documents for us.
For example, a <cite> tag around a book's title would tell the software that this was a title. It could render that part of the text in italics for human eyes, but also add the title to a list of works cited in the document. With the aid of HTML tags, the software could easily take care of all these clerical tasks for the human.
But Web authors didn't care about machine-readability. We wanted an enhanced experience. We wanted wizzy. We perverted the HTML tags, making them from logical tags into physical ones. We started using <i> (italic) instead of <cite> because we liked the way it looked. The result was that we had more control over our pages' appearance. The side effect was that software could no longer understand our documents.
We ended up with the worst of both worlds instead of the best. We have HTML, which is an insufficent physical markup language. But by misusing HTML, we have also made it into an inadequate logical markup language. In other words, it's not suitable for any purpose.
HTTP and HTML were so much better than the previous standards (ftp, Gopher, Telnet, etc) that the Web immediately overwhelmed the market and drowned out all other traffic on the 'Net. It was so much better that it didn't have to be compatible with previous technologies. It just took over. I don't think we'll see that kind of revolution again anytime soon. In other words, any new standard must be backward-compatible with HTTP/HTML if it is to be accepted.
I visited Trellix's Web site. It's not clear to me if they plan to support HTTP/HTML. But if they want to be a big player, I think they have to. Unfortunately, if they do, they risk being perceived as just another Web browser. If they get off the ground, though, it sounds like their product can put back into the Web a lot of what was lost when we misused HTML for wizzy purposes. I hope it works!
Another thing that the 'Net has shown us is that if we agree on a language (TCP/IP), we need not agree on an operating system. There are email and Web applications for every conceivable platform, and (generally) you're free to use any platform you want, as long as you're willing to speak TCP/IP. Unfortunately for Trellix, they seem to only have a Windows version right now. In addition to HTTP/HTML compatibility, they need to support at least Macintosh and Unix if they're to be accepted on the 'Net. If they want their software to be used, it's a lot easier for them to port it to other platforms than for the rest of the world to go out and buy Wintel machines so that they can run Trellix.
I hope what I've said here makes sense. In fact, I hope it makes enough sense for you to share it with Mr. Bricklin. The Web needs Trellix, so I hope it is a success. However, the Web is too large and "just good enough" to hold its own unless Trellix is willing to fight the battle on the Web's terms.
I loved Trellix. I think we're looking at the way of word processing for the web. I don't know what to call it yet (Dynamic Documents? Web Processing?) -- maybe you'll come up with a category name?
From: amy@home.cynet.net (Amy Wohl);
Sent at 7/22/97; 4:46:42 PM;
Re:Challenging HTMLI'm writing my article on this tomorrow. I'll send it along. But you're absolutely right. This is a needed step in the right direction. I can't wait to run, not walk, along what feels like just the right path.
XML (Extensible Markup Language) is the attempt to go beyond HTML. It is a simplification of the SGML (General Markup Language) for creating DTD (Document Type Definitions) of which HTML (you know that one) is one. Whew!
From: delza@voyager2.cns.ohiou.edu (Dethe Elza);
Sent at 7/22/97; 4:28:12 PM;
Re:Challenging HTMLXML has three parts: Structure, Style and Linking. Structure is the part HTML was intended for, meta-information about content, no rules for presentation. Style goes beyond what is presented with CSS to allow for true WYSIWYG (if that's what's desired, there are many reasons NOT to do WYSIWYG also). Linking also goes beyond what HTML offers, with robust linking which is consistent and extensible.
XML won't get us entirely away from plug-ins, but it could make them invisible. The "browser" will be able to parse the XML and validate it, then get the plug-in to display it, if neccesary. I know you're not crazy about Java, but this becomes fairly easy with Java, especially if you use the CORBA object model for distributing the components.
So far Microsoft has begun using XML to create it's "push" technology. We'll all be hearing a lot more about XML very soon. I'm using it in my master's thesis on distributed word processing. I'm also using Frontier as a model--I hope to talk with you about it at MacWorld or on-line.
One other note: I'm converting my web pages to Frontier and making decent progress. I'll let you know when they are more or less done. I've been waiting for summer to have some more time to work on it, now summer is half over and it hasn't gotten significantly less busy. Oh well.
How's the garden?
--Dethe
P.S. The XML spec is available from W3C at
http://www.textuality.com/sgml-erb/WD-xml.html
and there are various articles about it at Sun, Microsoft and WebMonkey.
1) In a perfect world, linking would be a core service and wouldn't be relegated to a user program like a browser. There are some researchers who are bouncing around the idea of generalizing hypertext into "structural computing" -- I believe there are papers at the Hypertext '97 conference and Peter Nuernberg (bush.cs.tamu.edu/~pnuern -- although I couldn't check the URL) is a leading proponent of the idea. Imagine -- everything can be linked: one metaphor for aliases, hypertext, directory browsers, info pages, toolbars, etc. When I helped run a workshop (WWWW4) on "Design and Engineering on the Web," we all agreed browsers would go away, and their functions would be dispersed to the operating system and applications. Instead, they seem to be sucking the rest of the computer into them.
From: Kenneth.J.Meltsner@jci.com (Ken Meltsner);
Sent at 7/22/97; 2:21:49 PM;
7/21/97 DaveNet2) HTML is a rude hack. Luckily, HTML 4.0 should be the last of the bunch. We need to support the XML folks so we can deliver content without embedding all the presentation bits, and we need to find a good presentation format (PDF? Flash? ?) for all the presentation-heavy pages. Add Java and a scripting language to the mix and we can stop overloading poor little HTML.
3) Where are our group capabilities? Berners-Lee was dead-on for the most part (I came across his original research proposal for the Web), but he promised to get services like annotations, recommendation lists, etc. working so the information channel could be two-way. Right now, most Web sites are more like magazines or TV programs than communities of practice or workgroups.
Dan Bricklin is a real smart guy, but I hope his tool can talk XML. [If you need more info, I can drop you a Powerpoint talk I'm putting together on Web/Intranet services and standards.]
Thank you for finally stating this clearly, and for proposing alternative solutions! People always ask me why I won't use "WYSIWYG" HTML editors...
From: beloy@kinkaid.org (Ben Eloy);
Sent at 7/22/97; 2:10:12 PM;
Re:Challenging HTMLMost professionally designed sites are done with emacs.
From: toxic@wired.com (Jeff Burchell);
Sent at 7/22/97; 10:57:50 AM;
Re:Challenging HTMLNotepad, Dave? Come on.