Up and Down the Ladder of Abstraction

http://worrydream.com/LadderOfAbstraction/

by macrael • 15 years ago

733 72 15 years ago

macrael• 15 years ago

Definitely read the appendix "Tools & Implementation". He expresses frustration with the current state of the art in building visualizations.

My guess is that he is using his javascript library for "reactive documents" called Tangle http://worrydream.com/Tangle for all the demos in this article.

I'll just add that I really find this whole piece astounding and many levels above the majority of things I find on the internet.

funkah• 15 years ago

Yes, "astounding" is the word.

keithpeter• 15 years ago

Excellent stuff, as others have mentioned, I thought of Tufte meets Web.

The "Tools & Implementation" section had me thinking of the pure data sound programming environment - the way you drag the numerical parameters in the code (cheekily reminiscent of processing) that draws the circles to change values.

tangle.js is referenced in the html of the page, and I shall certainly be looking into that for maths visualisations at a more humble level.

yariang• 15 years ago

I think the best part of it is the car at the top. I can easily play with that for a few minutes. They do say 80% of the gains come from 20% of the effort right? :)

ineedtosleep• 15 years ago

Nice little bit up there, but it was absolutely frustrating at first since I wasn't able to go down the page with my up and down keys and I was wondering why for almost 5 minutes.

arkitaip• 15 years ago

Fascinating read and absolutely gorgeous to look at. This is the guy behind Magic Ink http://worrydream.com/#!/MagicInk

kragen7• 15 years ago

I agree. And I think you can see both this essay and Tangle as further work in the direction he identified in Magic Ink.

asolove• 15 years ago

Totally speechless. Feel like this is what I ought to be working on/with. I look forward to the day when online news sources take advantage of the medium to cover wars and budget proposals in this format.

danso4• 15 years ago

Wow.

This is simply one of the best guides I've seen to anything...ever. It's uncommon for an online guide to have well-written text, even rarer for it to leverage HTML's structure and organization, and almost unheard of to incorporate interactive useful elements that gel with the textual sections.

Well done.

danso4• 15 years ago

Looking through the rest of the author's (Bret Victor) website, I'm in awe. If this guy wanted to declare himself the Edward Tufte of interactive visualization, I think he'd have a good case for it.

egiva• 15 years ago

Yeah, and taking a look at his CV he also did a lot of Apple-interface design work. And has a best selling product (neat synthesizer). And a lot of other top-notch work. This guy is a definitely a neo-Tufte candidate for the digital era.

wiredfool• 15 years ago

This is beautiful.

jonnycowboy• 15 years ago

Is this behaviour supported (and easy to use) on mobile devices as well?

A great use of this would be for electronic math textbooks for showing students (high school) how various parameters influence graphs, etc.

xenonite• 15 years ago

Impressed on iPad here...

danso4• 15 years ago

From his CV: http://worrydream.com/cv/bret_victor_resume.pdf

Apple: >>I designed the initial user interface concepts for iPad, iPod Nano, and half a dozen experimental hardware platforms. Initiated, designed, and prototyped over seventy concept projects, including radically reinvented interfaces for video editing, animation, drawing, learning, collaboration, mail, photos, and much more. Invented features for Mac OS X Lion. Worked with designers and engineers from all parts of Apple. Routinely presented to top-level management

2bit• 15 years ago

Crashes on iPad 1 for me :(

Wilduck• 15 years ago

I was very surprised how well this worked on my android phone. Normally these sorts of things are frustratingly worthless. Here I was able to interact almost as well on my phone as on my laptop.

Color me impressed.

enjalot2• 15 years ago

This is the future of education. He just laid the blueprints.

VBprogrammer• 15 years ago

I don't know whether this was a one off, a subject which he was able to knock out of the park or if you are right. I do think someone should pay him a lot of money to find out.

icandoitbetter3• 15 years ago

Definitely not a one-off project. It's part of his "Kill Math" series that he started earlier this year. Check out the introduction: http://worrydream.com/KillMath/

enjalot2• 15 years ago

regardless of whether he continues (it looks like and I hope he will) my point was rather that everyone should follow these blueprints.

I know I'm inspired!

colanderman• 15 years ago

Beautiful site. But I hope anyone actually trying to solve the example problem of guiding a car on a road uses a PID controller and not a binary threshold. Here's an interactive example for those interested: http://fstutoring.com/youlearn/pid/controller

joshhart• 15 years ago

I agree. It's a nice site and a lovely visualization, but it's not the right way to solve the problem. A PID controller or something similar like a Kalman filter is the way to go.

The Anybots robots are a great example of a PID filter in action.

icandoitbetter3• 15 years ago

Bret Victor is one of the great people working in computing (not computer science) today. I expect something paradigm-shifting from him in the coming years.

div• 15 years ago

I'd never come across his site or anything by his hand before, and feel like I've missed out on a lot.

On the upside, I'm excited at going through all of his archives. I haven't felt inspired like this in quite a while.

danso4• 15 years ago

You feel exactly the same way I do.

kragen7• 15 years ago

This is a good example of how to make a high-confidence prediction about the future: look for something that's already happened, and predict that it will have happened soon.

In other words: he's already published paradigm-shifting things!

icandoitbetter3• 15 years ago

Obviously hidden in my statement was the assumption that what he's produced so far is not paradigm-shifting stuff. But go ahead, be snarky.

tambourine_man• 15 years ago

Damn this guy is good. Check the whole site out, well worth it.

xtacy• 15 years ago

Very impressive, but I am not sure his comment about Mathematica is right. The Wolfram Demonstrations project is about aiding such explorations in Mathematica. It may not look as pretty, but it's definitely capable of enabling explorations using the Manipulate[] command.

http://demonstrations.wolfram.com/

Example: http://demonstrations.wolfram.com/DigitalTiltShiftPhotograph...

http://reference.wolfram.com/mathematica/ref/Manipulate.html

programnature2• 15 years ago

Work is very nice by web standards, but Mathematica's version of interactive documents is much more sophisticated. Strange that he seems to be unaware of what Mathematica can do, given the rant on tools at the end. Its pretty much exactly what he wants, only just proprietary and not on the web.

bbq2• 15 years ago

"Its pretty much exactly what he wants, only just proprietary and not on the web."

Which is exactly what he doesn't want. The authoring tools should be accessible to everyone.

Secondly, after exploring his site I'm guessing Bret would reject Mathematica even if it were free & open source. Because even then it's not accessible to everyone - you still need to know how to program.

nagrom3• 15 years ago

I'm a physicist studying nucleon structure. This is a field so complicated that other physicists typically shake their heads and mutter "must be mad". We typically work at several layers of abstraction.

This (Tangle) is the one tool that I've seen that is capable of communicating well to non-physicists what I do, without their needing to know Mathematica or C++ in detail. I really need to get around to writing a web page that does that. Awesome. Inspiring. Ridiculously, breathtakingly good. Bret's entire website is outstanding.

kragen7• 15 years ago

You mean you've used Tangle for that, and it worked? Are your explanations public?

nagrom3• 15 years ago

Work in progress; if/when it gets finished, I'll post it on here.

rottencupcakes2• 15 years ago

Does it bother anybody else that he fails so miserably at solving the example he uses for this article: keeping the car on the road?

nagrom3• 15 years ago

No. He's not trying to keep a car on the road. He's trying to let you understand the problems behind keeping a car on the road, so that you should solve that problem for yourself.

rottencupcakes2• 15 years ago

I'm sorry then. I just personally find that how dumb his example is keeps this from being interesting at all.

Why not pick an example where his clear ability to make useful visualizations reveals some startling insight about an initially complex system that he then abstracts away to simplicity?

EDIT: Thanks for downvoting guys. If you felt this blog post revealed something fundamental and astonishing, how about explaining it to me instead of downvoting me out of anger? I think I'm done posting on this website.

slewis• 15 years ago

No. His point is not about keeping a car on a road. Read the "Iteration" section.

rsanheim• 15 years ago

Actually, the opposite. As I read through the examples and played with the demos, several ideas for ways I would change the algorithm came up as I saw how his simplest algorithm failed. Which I think was precisely his point. Once you have these tools in place and can move up and down the ladder, you have learned much more about the problem domain, and better solutions become apparent.

michaelchisari• 15 years ago

This is definitely one of those moments where something inspires me to look at everything I'm working on and want to throw it all away, and rebuild it at this level.

gfodor• 15 years ago

Bret Victor, as usual, is a humbling reminder of how much I don't know and how constrained my thinking is. Bravo!

AdamTReineke• 15 years ago

That was fantastic. In case you missed it, use the arrow keys to move the car in the header to jump up the ladder.

pyre• 15 years ago

Annoyingly, the arrow keys are hijacked, and I can't use them to scroll the page.

phzbOx2• 15 years ago

Looking at his website, his resume and his project, I just feel like stopping everything I do. A little bit like if you start to run for a couple miles and after 5-6 when you're pretty tired, you ask your friend how much remains. And he smile at you, start running even faster and says 195miles.

Or, it's like playing Starcraft agains a good player. You get beaten pretty hard but you still played your best and are proud. But then, somehow, you play a real professional gamer.. and you just feel like stopping playing that game. Nothing works, you look like a total beginner, you get 5/0-ed, and then, you learn he was on the phone the whole time.

Meh.

sriramk• 15 years ago

this is exactly how I feel after spending the last 60 minutes reading through his site.

esk• 15 years ago

If someone had told me his work had been done by a devoted team of a dozen people, I'd still be tremendously impressed that only twelve people managed to accomplish so much. The fact that it's one man? It's chilling.

Try looking at things another way—he's generously provided us all with an incredible foundation upon which to build our own things. After having read a few of his essays, I certainly feel changed.

phzbOx2• 15 years ago

Ok well, it's been 3 hours I'm on the website; reading, looking the source, bookmarking lots of things, buying new games and books. This guy is simply amazing. Please, do yourself a favor and get a look at these links: http://worrydream.com/#!/Links

kragen7• 15 years ago

I feel exactly the opposite way. I find this essay and his site in general really inspiring, and I always get lots and lots of ideas for things to do when I read them! I'm also really glad he's implemented Tangle, because it means I don't have to; I can just use it. Although I haven't yet.

jonmc122• 15 years ago

Exactly - the takeaway is that Bret can create amazing things as a result of building off of his own creations. He's not running, he's building reusable components for a rocket ship.

sireat• 15 years ago

Add Bret Victor to the list of people to read about when you want to be depressed about your lot in life. This list includes Fabrice Bellard et al.

Bret is even more impressive since he combines hardware, software and UI skills.

Why depressive? As Neal Stephenson mentions upon Hiro meeting Raven in Snow Crash. If you are over age of 25, you realize that there are certain levels of excellence that you will never reach even if you move to that Shaolin monastery.

I think some people have this innate drive to excellence (how much of that is acquired and how much is inherited is another discussion) and apply themselves with full dedication. You can see Bret's teenage projects were already pretty impressive.

Ever since then Bret's toolbox of skills has only multiplied till you see the effect today.

jpitz2• 15 years ago

I see what this author has done, and I wonder to myself if I could ever possibly have the time to write myself tools to do the things he suggests. Then I wonder if I have the time not to.

defdac• 15 years ago

The intro is a really good guide on how to build interesting (exploration) games. Minecraft comes to mind, for example people making charts of ore beta distributions in effort of understanding the inner workings of the world they play.

jsilence• 15 years ago

Awsome!

nitrogen• 15 years ago

The only thing that could make this better is if the arrow keys still worked for scrolling (Firefox 7).

Up and Down the Ladder of Abstraction

Related Stories