Douglas Engelbart's HyperScope 1.0 Launched

ReadWriteWeb writes, "HyperScope 1.0 is a new Web app based on Douglas Engelbart's 1968 NLS/Augment (oNLine System). Engelbart and team have been working on Hyperscope since March of this year in a project funded by the National Science Foundation. Its aim is to rebuild portions of Engelbart's NLS, on the Web, using current Web technologies such as Ajax and DHTML. In effect it gives an advanced browsing experience, including classic hypertext features like indirect links and transclusions of remote pieces of other documents. HyperScope has been completely built with open source JavaScript toolkit Dojo — meaning that everything is done on the client-side."

is this really a good idea?

[Desolator144]

If it takes you 10x longer to design the content and the person viewing the page can find what they want 10x faster then is it really a net gain? (no pun intended) It still is a good way to organize information and beats the crap out of the way some webpage makers do it. It sure would save surfing/researching time but is that a good tradeoff for a massive slowdown in content creation speed because of the page maker having to add all the meta data type stuff?

Re:is this really a good idea?

[PhrostyMcByte]

If it takes you 10x longer to design the content and the person viewing the page can find what they want 10x faster then is it really a net gain?

Maybe it won't make sense if only a single person is viewing it, but with hundreds, the gain could be very useful. Not just to the readers, but the writer too - if 100 people view the doc and 75% of them are able to understand it better thanks to this, that's much less questions the writer will have to waste time answering.

Re:

[Qzukk]

If it takes you 10x longer to design the content and the person viewing the page can find what they want 10x faster then is it really a net gain?

What if two people find what they want 10x faster? (Other than what would really happen: companies would switch from 10 words on an ad-filled page to 5 words each?) There are generally more content consumers than creators, both in the general sense as well as in the case of just about any field.

Re:

[RuBLed]

and it seems to be not compatible with Opera [the browser I'm using now] so I tried loading it in IE7 RC1 and only got it to work after around 4 or 5 "permission to run active x controls" (you know, that yellow information toolbar)

So then I browsed the demo, hmmm.. the interface looks fresh and promising, but for compatability reasons, it is much easier (and faster) to browse a document thru html links or by printing them. Although I believe this HyperScope concept (or something similar) is bound to be popu

Re:

[PhunkySchtuff]

If it takes you 10x longer to design it, and users can find what they're after 2x faster, and there are 5 users, then yes - there is a nett gain.
Development/design is generally done once, but reading and digesting the information happens over and over again, by many many different people.

Wiki?

[ScruffyNerf]

IMHO it seems that this would be a good add-in for community reference materials like wiki's.

Unfortunately the spammers will probably hook into this first... or the 'Web 2.0' crowd.

Re:

[ballermann]

Eugene Kim [1] (why didn't the poster give him credit? It's mostly his work, but I guess Doug is just a tid bit more popular) presented the system at WikiSym2006 [2] and did indeed show how to integrate this as additional view into a Wiki (I think he used his PurpleWiki [3])

[1] http://www.eekim.com/blog/ [eekim.com]
[2] http://wikisym.org/ws2006/ [wikisym.org]
[3] http://www.blueoxen.com/tools/purplewiki/ [blueoxen.com]

How does it compare to Xanadu?

[Beryllium Sphere(tm)]

Community orientation, granular targeting, indirect linking and of course transclusion were all part of the Xanadu vision. How's this different, except for the notable distinction of actually existing?

Re:

[fm6]

All the buzzwords are completely different!

Re:

[Eideewt]

Existing seems to be HyperScope's big advantage.

Re:

[Mateo_LeFou]

Also it doesn't seem to go in for micropayments everywhere.

Re:

[Myself]

Ahh, glad I looked for mentions of Project Xanadu [xanadu.com.au] before posting! Yes, this sounds like a rehash of the same ideas.

Existing is good.

Re:

[BradNeuberg]

This is also a reimplementation of the original hypertext system, the one that invented hyperlinks in the 60s, named Augment/NLS, so it predates Ted Nelson's Xanadu and his book Computer Lib. Just to note Ted Nelson actually invented the word "hyperlink" and "hypertext", but Douglas Engelbart had a working hypertext system first; he just called them 'addresses' and 'links'. HyperScope is a clone of parts of Augment for the contemporary web.

Actual link

[generic-man]

Actual site: http://hyperscope.org/ [hyperscope.org]

Blog blog blog blog blog, blog blogpost blog blog...

Re:

[tehcyder]

Blog blog blog blog blog, blog blogpost blog blog...

Your ideas are intriguing to me and I wish to subscribe to your newsletter.

AHA! caution: massive insight inside

[Desolator144]

I figured out how to make this super popular! With all these capabilities, they should just take that tiny extra step and have disaster scenarios like a monster slowly eating the info on the page and you have to stop him or a black hole at the bottom of the page sucking up all the text and scrolling you down faster and faster as it grows. That would be cool and would really spice up boring web documents. I think it would even encourage kids to read long research papers more :-)

So let me get this straight....

[supasam]

The demo I looked at on the page seemed to be to be an html doc, with the difference being that instead of clicking the words, you would awkwardly click the column to the left to the line in which the thing you want to link resides? I must be missing something cause I don't see how this might be any easier. True I didn't actually read the paper I was clicking through which happened to be on this topic, but how many web pages do I really actually read anyways? A little help?

Re:

[RuBLed]

If I'm right, it is just one file and to put it simply, it is just like reading an html document with the usual anchor tags, where "sections" (div?) appear and disappear as you read the document. If I'm still right, it would create an illusion that you are reading several pages but in reality, you only made that part of the page visible by clicking on the appropriate "links" or "arrows". The article says everything is client-side so I assume the document is all there.

Re:

[Anonymous Coward]

So it's basically the same as the "New Discussion System" which is currently being tested right here on Slashdot?

Re:

[HiThere]

Except that the Slashdot version works better, and is more intuitive without instruction.

Zigzag looked more promissing, but seems to have been dead since 2001, and moribund since 1999. As I read it, Zigzag appears to be based around a resizeable vector of doubly-linked lists that were conceptualized as being orthogonal. Could have been *very* interesting, but probably also quite confusing to navigate.

Re:So let me get this straight....

[TrappedByMyself]

Yeah, I really don't get it either. The little buttons appear even if there is no useful action to take. If the text is expanded, or whatever its doing, you can still click the little page icon. Doing so, it jumps to some other location, then jumps back to somewhere close to where you were, could be the same line, could be a few lines off. If you hit a few lines off, the technology has successfully disoriented you and has given you no way to get back to where you were before.

I guess if the team's goal was to create something which is even more of a navigation nightmare than Flash, well, then, They're Winner!

Re:

[supasam]

Indeed, it took me a couple (few) clicks to realize that not only was I still on the same page, but that the page I was on wasn't as long as I thought it was. If they'd have demo'ed something huge like the Sams FreeBSD 6 unleashed book (weighing in at about 877 pages including the index) or something like that, maybe there could have been something to it. I don't see this being all that useful to me as of yet.

It appears to be a web based outliner

[hey!]

Based on approximately 30 seconds clicking through links.

If so it's not a bad idea. If my impression is correct, then think for a moment about this idea:

        hyperscope : wiki :: outliner : word processor

There is no document you can produce in an outliner that cannot be produced in a word processor, whereas there are practically infinite documents you can create in a word processor. Which is what makes an outliner useful. Classification is such a basic and useful mental pattern, putting an amorphous blob of thoughs into the form of an electronic outline goes a lot of the way to organizing it into something coherent.

My experience with wikis is that if you have one really mentally disorganized person with time on his hands, he can quickly turn important parts of the wiki into mush.

There's hope.

[exp(pi*sqrt(163))]

Version 1.0 conceived in 1968, implemented in 2006. There's home for Duke Nukem Forever then.

Re:

[bobby1234]

but no HOPE for you.... sorry but nice typo

People are strange...

[pianoben]

I'm surprised that the man who spearheaded the development of the mouse would produce such an unnavigable system; you can't even scroll down with the arrow keys! Of course, in his vision, you're using a mouse and a 5-key chording keyboard; no third hand seems to be available for said arrow keys. But to not enable the scroll wheel, either?! No, thanks...

Re:

[civilizedINTENSITY]

Well my scroll wheel is sure working. So are my arrow keys. Did you try the original website's demo [hyperscope.org]? It looks like HTML merged with an Outliner.

Moo

[Chacham]

here [hyperscope.org]

For example, every paragraph in a HyperScope document has a location number, an address corresponding to the paragraph's location in a document. For example, the second paragraph in the top-level of a document has the location number 2. To link to this paragraph, you can use the address:

    http://foo/bar.opml#2 [foo]

To jump directly to the second paragraph while viewing a document, you can click on the Jump button, type 2, and press Apply.

Isn't this just an anchor tag on drugs? Just make it part of HTML 5.

Re:Moo

[Lars Arvestad]

Isn't this just an anchor tag on drugs? Just make it part of HTML 5.

I would actually like to see something like this in HTML. I have always found it a bit weak that you have to explicitly put anchors in your document to be able to jump there. Of course, I have not been clever enough to suggest an improvement or even voice my disapproval.

Some way of saying "fifth P element of this document" or "H1 element with text 'The Main Point'" should all be pretty easy to implement for a client program.

Re:More Moo

[Sir Runcible Spoon]

I always wanted to link to arbitary points in other people's documents, perhaps to a paragraph containing a particular piece of text. So I guess it would be xpath expressions in links that we need.

Re:

[Fred_A]

I suppose this could be done with some sort of JavaScript trickery (a kind of linkTo(URL, RegexKindOfThing)).

It requires some cogitation and a bit of coffee though. A general purpose function for this would indeed be a very nice tool. It should be part of the Dojo toolkit [dojotoolkit.org] or something like that...

Re:

[Sir Runcible Spoon]

I was thinking of this as requiring an extension to HTML. Doing it using existing JavaScript technology would be tricky as it would mean running JavaScript inside the target page. Even if you find a way to do this, you cannot be sure that someone at Mozilla or Microsoft woun't see it as a security problem and close the hole.

I don't think it would be possible with AJAX unless the target document is on the same server, this is a restriction of XMLHttpRequest [microsoft.com]. (This where I would like to link to the "Securi

Re:

[Chacham]

I have always found it a bit weak that you have to explicitly put anchors in your document to be able to jump there.

Then get a tool to put the tags in for you. Don't mess with the standard that allows the granularity of actually defining it.

Re:

[kthejoker]

Exactly, just write your page in XML and then transform it appropriately with XSL.

if (element in anchorList) {add anchor before element}

is a fairly simple conditional to apply.

Or add it to your processor as a function. But parents is right, granularity is the key to all good HTML specs.

Re:

[Lars Arvestad]

Yeah, sure, but what if it is not my document? If it is a document I own, I wouldn't even bother with using anything other than my favourite authoring tool and put the anchors that I need where I want them. That is probably easier than figuring out what part of the document I want to link to. And if I use some auto-generated anchor using XML+XSL as someone suggested will just cause the pain of figuring our how a new edit would affect the auto-generated anchor-ids.

The point here is that I want to link to a

Re:

[Chacham]

This comment: If it is a document I own, I wouldn't even bother with using anything other than my favourite authoring tool and put the anchors that I need where I want them.

Previous comment: I have always found it a bit weak that you have to explicitly put anchors in your document to be able to jump there.

What i commented on, and what you are replying to seem to be different things.

Re:

[Lars Arvestad]

Good point. What I had in mind in my first post was: ... explicitly put anchors in a document ... rather than ... in your document...

Sorry for the confusion.

Re:

[Chacham]

Ah OK, at least i understand the sentiment now. :)

Re:

[zacronos]

That only works if you are linking to webpages you create or can edit. If I'm linking to someone else's massive webpage with no anchor tags, it would be nice to be able to have the link automatically take them to the paragraph/sentence/image I want (just like an anchor tag allows, but without the anchor being explicitly coded). The alternative is to give the link, and say "scroll down to the 5th paragraph in the section on asshats, next-to-last sentence".

GP's suggestion wouldn't have to change the way

Re:

[tehcyder]

I have not been clever enough to suggest an improvement or even voice my disapproval.

You must be new here.

Re:Moo

[hey!]

http://foo/bar.opml#2

I think it has to do with how the reference is generated.

OPML is an XML standard for creating outlines. You use your favorite OPML outliner, dump it on the web, and suddenly you have a web version of the outline. OPML was not designed as a web standard; once you put an OPML document on the web, it immediately becomes obvious that every node in the outline ought to have some kind of address you can link to.

I'm not terribly thrilled by this approach, but something like it is necessary. I'd rather extend OPML so that every node has a UUID, as well as an optional human readable idea. That way you could cite the node as http://foo.bar.opml#dbcb7190-3d9b-11db-a98b-080020 0c9a66, even if the author did not think it worthy of a human generated ID. The advantage of this system would be that citations would be stable, even if users reorganize the outline. Which is kind of the point of outlining.

Re:

[Chacham]

OPML was not designed as a web standard

Neither was HTML.

rather extend OPML so that every node has a UUID, as well as an optional human readable idea. That way you could cite the node as http://foo.bar.opml/#dbcb7190-3d9b-11db-a98b-08002 0 [bar.opml] 0c9a66, even if the author did not think it worthy of a human generated ID.

Then just add character or paragraph count. Using current regex expressions, it really should be that hard to implement.

The advantage of this system would be that citations would be stable, even if

Re:

[hey!]

Using current regex expressions, it really should be that hard to implement.

Not if a paragraph was edited.

A simple hash could do that.

Not if the paragraph was edited. hash("I like hashes because their good for identifying paragraphs") hash ("I like hashes because they're good for identifying paragraphs")

Of course you could use some kind of serial genrator, but what if you cut paragraph from one outline and put it into another? UUIDs are just as easy to use and much better. You can move a paragraph fro

Re:

[Chacham]

You can move a paragraph from one site to another without it losing its identity.

But you said earlier "That way you could cite the node as http://foo.bar.opml/#dbcb7190-3d9b-11db-a98b-08002 0 [bar.opml] 0c9a66, even if the author did not think it worthy of a human generated ID."

Well, if the author didn't give it an id, and the id was autogenerated, and then the author changed some of the text, what good is the id? Would not the author be required to decide if the change was major or minor, having a minor change keep t

Re:

[hey!]

You are assuming that all documents are essentially unstructured.

For truly unstructured documents, such as short story or a blog post, you are right.

However, there are other kinds of documents, particularly outline structured ones, where nodes in the text do play an indentifiable function. For example, I could say section 1.5.2.1 of our outline, and its descendents, list all the known regulatory requirements our project has to comply with. Or section 2.1 provides a summary of our product position versus

Re:

[Chacham]

I hear the point, i just don't see this as a breakthrough or anything, and certainly not something thayt isn't mostly alreayd done.

Re:

[KermodeBear]

Then just add character or paragraph count. Using current regex expressions, it really should be that hard to implement.

This reminds me of one of my all time favorite programmer quotations, by Jamie Zawinski:
Some people, when confronted with a problem, think "I know, I'll use regular expressions." Now they have two problems.

Re:

[Chacham]

And you just reminded me of something i heard during an oral test. It went something like this.

The tester asked a question, the student gave an elaborate answer. After he finished, the tester said "That's an excellent answer, but for the wrong question."

More on HyperScope OPML From The Coder

[BradNeuberg]

Hi there; I'm the software architect and implementor of HyperScope, and thought I'd provide some more info on the OPML capabilities of what you can do with it.

HyperScope is like OPML and HTML hyperlinks on steroids. There are so many different kinds of addressing your nodes that its not even funny. You can do direct addressing using hiearchical placement, even if there are no anchors:

mydoc.opml#2A

Where 2A is the second node (2), followed by it's first child (A). Node numbering always alternates between lett

Now I'll never get to sleep

[MarkusQ]

Great, now I'll never get to sleep.

First FreeDos 1.0, and now this. If I go to bed now, I'm sure to miss the story I'm really waiting for; just think of it, Babbage's Analytical Engine, completed at last! Will it work the way he thought?

The suspense is killing me.

--MarkusQ

Re:

[James Youngman]

yes [fourmilab.ch]

Re:

[velorg]

Where you been sleeping, man?
That hardware (=crucial, number-crunching part of the Mill) was finally implemented waaaay back in 2000 or thereabouts, and is now on display AND running (crank-powered) in The Science Museum in London, UK. Scientific American had a feature on it around then,too. Sorry, no urls.
Along the way, the two "implementers" discovered several "mechanical" errors in Babbage's original drawings, which would have prevented the Engine from fuctioning even had it been built; so they corre

Yes, I know

[MarkusQ]

Yes, I know.

My post was a test of a new form of communication that we've been trying out over here (called a "joke" or "jest"). It aims to amuse, but at the occasional cost of some technical accuracy.

Hope that helps.

--MarkusQ

virtual machine

[drDugan]

hot damn! browsers really will be a virtual machine.

I give us about 4.9 years until self-aware computers emerge

4.9 years?!!!?

[headkase]

:0
As a computer, I find your faith in technology amusing.
;)

Enough with the snide remarks....

[FlyingGuy]

Ohh wait this IS> /. after all...

Now here's something we hope you'll really like!

Its seems to me that its time for a major shakeup in HTML. While CSS and other things make bold attempts at making the WWW usefull for something other then very basic page display it does seem rather lacking.

Page inclusion based on simple tests! The machinations one has to go through to make what seems like an obvious feature like if(X=Y,this.html,that.html) actualy happen require scripting and all sorts of PHP like occurances, java, javascript and what have you.

CSS gets ever more complex to simply accomplish a text area on the rendered page.

So now the idea HyperScope gets trotted out from a dust filled closet. I can't really see how that is going to help matters rather then just muddy the waters even more.

Isn't it really time for a MAJOR overhaul of HTML and the HTTP protocol? HTML as it was created was sufficient for the basic exchange of information, formatting was limited and things worked fairly well. Microsoft came along and realized that this could be taken MUCH farther and started doing things their own way, which was good for them, not so good for everyone else and was properly derided for doing so in a uncooperative way. But was it really? I agree that MS did what they did out of pure self interest, but did they suggest these things to various committies only to have them shot down because it was MS centric only? I don't know the answer.

I for one think its time to just scratch HTML as it currently exists and scratch CSS right along with it, and come up with something that actualy works smoothly and works as designed!

Basic fundemental layout strategies like Areas, Menus, Images, scrolling, layers and all the primitives sould be part and parsle of an HTML-Like specifcation, not an add-on cludge like CSS. The specification for this should rightly be decided upon by the major browser venders and fully supported by all the web server vendors. It should be code modules written to the specification that are freely available to anyone who wants to link them into their rendering engine. Further more, they should be written in ANSI C and all supporting libraries are included also written in ANSI C.

Re:

[FlyingGuy]

Ahh yes, but it IS broken, thats the problem

Re:

[wralias]

I for one think its time to just scratch HTML as it currently exists and scratch CSS right along with it, and come up with something that actualy works smoothly and works as designed!

I'll grant you that, yes, there are some limitations and frustrations with CSS. But to scrap it entirely is nonsense - why start over from scratch, when CSS 2 and 3 are actually steps in the right direction? The newer specifications for CSS contain much more power and control over layout and style. The problem is not with

Re:

[FlyingGuy]

Imagine this...

Document.Background = white

Document.Styles = /styles/PageStyles.stl

DefArea('TopHeader',(0,0),100px,Document.Width)

DefMenu('MainMenu','/menus/main',TopHeader)

TopHeader.ApplyStyle = Document.HeaderStyle

MainMenu.ApplyStyle = Document.MainMenuStyle

Or something like that.

Why don't do the right thing anyway?

[master_p]

The right thing would be to forget about HTML alltogether and make a distributed platform where the browser IS the virtual machine, programs are lazilly downloaded like web pages and the GUI is a library, rather than a set of specs specified by a commitee.

Simple informational web pages are a subset of an application anyway.

Re:

[FlyingGuy]

I like your notion...

But how do you keep something like this from becomming a large bandwidth burden? Would it be every interactive site then has to be downloaded as an entire program? What you seem to be describing was MS's vision of ActiveX wouldn't you agree?

This became something that has caused probelms, security holes, etc.

The most tedious and problematic things like menu's that dont quite look correct from browser to browser unless some of the most increadably convoluted CSS is applied should be,

Re:

[master_p]

Bandwidth would not be a problem...first of all, the programs will only have to be downloaded once, or if they are updated. Secondly, the program will be downloaded lazilly, in compressed/tokenized form.

As for security, the code will be executed in a sandbox, like javascript.

Re:

[FlyingGuy]

Master_P,

So a defined menu system drop down, fly-out, combinations of those forms, downloaded once. It would encompas a raw menu system that could then be styled by a web page.

Style would entail things like tabs with rounded corners, square corners, color, depth, parent ( local to an equivilent of a DIV ) and such?

Lets start a project!

Re:

[master_p]

I would definitely be interested in a project, but from your comment it seems your understanding is different than mine.

First and foremost, I am talking about a totally new platform, not based on existing standards and web browsers.

My idea of 'platform' entails a virtual machine, a programming language, a 'browser' window, and a set of libraries.

The virtual machine would be distributed, i.e. it is responsible for:

a) interpreting the bytecode and caching its native representation both as a client and as a se

Re:

[Harry Coin]

Capital idea! Somebody rewrite decades worth of technology for me, but design it better this time! Also, make sure that all competing software companies cooperate on fully supporting it. Also, ensure that it mixes presentation information with content. Plus, was Microsoft really doing the wrong thing by extending HTML?

In addition, please hurry, because I am a very busy person and I don't like waiting.

The 1968 Demo

[mrdrivel]

In 1968 Douglas Engelbart gave a presentation of the NLS/Augment system in San Francisco. It's quite amazing. It uses a three button mouse and pointer (called a bug). It also uses several buttons operated by the non-mouse hand in a chorded fashion to select and copy text.

Video conferencing, group collaboration, the hierarchical presentation and hiding of data, spatial representation of data, hyperlinking are all shown in the demo.

There is a Quicktime of it here: http://www.invisiblerevolution.net/video-68-large. html [invisiblerevolution.net]

I first saw this in one of my computer science classes at Berkeley; we were all surprised at how much of what we think of as recent technology (last 10 years or so) actually existed in 1968.

NOTE: The video is rather long. The impatient (read: Slashdotters) might want to fast forward through parts.

silly made-up word

[DreadfulGrape]

Dictionary.com says

"No results found for transclusion. "

What I want to see is his presentation template...

[jpellino]

He gave a talk maybe 10 years ago here in CT, and his presentation was built on a template that essentially had a border of thumbnails around the outside like the ads on a restaurant placemat, starting at 12:00 and proceeding clockwise - each thumbnail either telltaled or swelled (like the OSX dock - can't recall) the cool thing was you had random access to your presentation, and you could jump to any slide from a single screen... you could build it from scratch, but I;m betting it was automated somehow.

Duke Nukem Nothing!

[WED Fan]

Wow, 1968? 38 years from concept to implementation? And idiots were giving the FreeDOS [slashdot.org] folks a hard time for their release schedule. :)

good alternative

[eegreg]

tiddlywiki is better and more powerful through the use of plugins http://www.tiddlywiki.com/ [tiddlywiki.com]