« Wiley blinks. | Main | A pause, because my brain is full and mushy »

Eric: Cause, then effect.

(From Nukees!. Click on the thumbnail for full sized new media perceptions!)

There is a core reality involved in this absurd joke. In a weird way, it reminds me of Reverse Polish Notation. For those of you who didn't go to college in the 80's (do Engineering calculators still use RPN today?), HP's high end calculators used a different, more efficient method of inputting numbers and getting out results, called Reverse Polish Notation. (No, it wasn't a Polish joke.) In Reverse Polish Notation, you actually input your information the way the calculator processed it, instead of making the calculator convert to it. So, if you wanted to solve for (2+3) x 6, instead of pressing "(" "2" "+" "3" ")" "x" "6" "=" and reading the answer, you'd press "2" "3" "6" "+" "x" with an enter key in between the numbers.

Why do this? I have no Earthly idea. But it makes sense to Engineers, and Engineers make the cars I drive and the electricity I consume in the computers and Tivos they made for me, and launched the satellite that gives my TV. So if Engineers like it, I'm all for it, baby!

My point is, we've adapted our interface to better suit the web and use it more efficiently. There's something terribly counterintuitive about blogs -- we're used to reading the old entries at the top and working our way to the new entries, in traditional media. To have the new entries on the top and scroll down to move back in time seems desperately wrong, when you're first getting used to it. But it makes vastly more sense for the web -- when I update Websnark, why shouldn't the newest entries go to the top of the page -- the bit that appears right in the window when someone shows up to read it. This way, people don't have to scroll to the bottom to see if I've updated or to read something new. It's always presented right where we want it -- at the top.

Well, sooner or later, going back and reading traditionally laid out information's going to seem a little screwy. Almost as screwy as using a standard calculator after you get used to RPN. And so this strip makes me a happy person.

Tasty, 2, Biscuit, Darren Bleuel, x + +

Posted by Eric Burns-White at December 15, 2004 4:26 PM


Comment from: baf posted at December 15, 2004 5:03 PM

Actually, 2 3 6 + x would be (6+3)x2.

Comment from: Pooga posted at December 15, 2004 5:16 PM

Since baf caught the error in polish notation (if I remember correctly, it would be "2","3","+","6","x" for the original equation), I'll just add that although I'm used to the blog format, after a fashion, I still find it somewhat counter intuitive. Usually, if multiple entries have been made since I last visited, I will scroll down to the last entry I read and work backward.

With Websnark it's not so bad, but several blogs (especially political blogs for some reason) have a tendency to refer back to a previous entry with an update or additional thought. The only way to make sense of the current post is to have read one halfway down the page. Or even off the main page with some more active sites.

Like I said, I'm used to it, but I still find it counter-intuitive to making sense of what you're reading.

Comment from: Ghastly posted at December 15, 2004 6:13 PM

Oh man. RPN. Does that take me back. When I was in the artillary we used HP-41 calculators in the command post for calculating the gun positions, elevations, angles, etc. etc. needed to lob an artillery shell onto a target a dozen or so kilometres away. It had a magnetic strip reader and then a book of a couple dozen little magnetic strips that had to be fed through the reader one after another in a tedious process to load the complex artillery program into the calculator.

I remember thinking "Geeze, if only there was a way to make my Commodore 64 portable. I could write not only a much easier to use program, but an easier to load one too". The closest thing I'd seen to a laptop in those days was the collosal Osborne portable CP/M computer which was the size of a big suitcase, weighed only half a tonne, and still needed to be plugged into 120v to function.

I'm almost afraid to find out if the Royal Canadian Artillery is still using the HP-41 for gun calculations.

Still, it was somewhat faster than doing the calculations manually on paper. I lobbed far fewer shells onto the Trans Canada Highway using the HP with it's RPN than I did using paper and pencil. And nearly obliterated one less church.

Good ol' RPN.

Comment from: Darth Paradox posted at December 15, 2004 6:25 PM

Along with baf's correction, I should note that "((Tasty x 2) + Biscuit) + Darren)" would be "Tasty 2 x Biscuit + Darren +", which doesn't look quite as outlandish. (Your formulation would yield "Tasty + (2 + (Biscuit x Darren Bleuel))".)

The general point is stack-based processing. The "Reverse" distinguishes it from how we normally read things, but it's actually read left-to-right. You push numbers onto the stack, and every time you push an operation onto the stack, you pop off the top two (or however many the operation takes) operands, apply the operation to them, and push the result back on. So "2 3 6 + x" would be interpreted in five steps:

2: push 2 onto the stack. Stack: 2

3: push 3 onto the stack. Stack: 2 3

6: push 6 onto the stack. Stack: 2 3 6

+: pop two operands, add, push result. Stack: 2 9

x: pop two operands, multiply, push result. Stack: 18.

And we're out of stuff, so there's our answer.

It's not just sticking all the operands at the end. The order in which the operations are placed among the operands actually dictates the order in which they're supposed to be evaluated (which we usually need parentheses for). Note that in both (2+3) x 6 and 6 x (2+3) we evaluate the + first (because of the parentheses), and in RPN those expressions are "2 3 + 6 x" and "6 2 3 + x" - in both cases, the + comes first, and it's supposed to be evaluated first.

I recognize that was probably more than you needed to know, but there you go. (I'm a computer programmer and actually learned this stuff in school...)

Comment from: Brandon E. posted at December 15, 2004 7:22 PM

I think if we're thinking in relation to blogs, it's better to think of printed materials as one "post." Then it doesn't really clash with how blogs work.

Comment from: Dave Van Domelen posted at December 15, 2004 7:59 PM

One of the big advantages of RPN is also that it makes programming simpler. It's in the stack, operate on it. And I can still revert to RPN after only a few seconds of adjustment, over a decade since I last used my HP-28S on a regular basis.

Comment from: Dragonshark posted at December 15, 2004 9:04 PM

I like newest stuff first, but only if they quote what they're referring to.

For many, taking newest stuff first predates blogging, if you used one of the major legal research databases (LexisNexis and Westlaw) you'd get your cases in reverse chronological order

newest stuff first, even if the seminal or most relevant materials were from years ago. (Although Lexis has a feature that lets you go by relevance, not sure about West, but the default is to the classic reverse chronological order)

News articles on those databases were also reverse chronological order.

of course, it could be worse, some discussion forums only allow you to read 1-5 posts at a shot, so if you take a look at one, you may have no idea what it's referring to (for example, the subconversations on a 100,000 post thread)

I miss law school, free access to research whatever random, sleepdeprived, caffeine addled thought might come to mind

Comment from: nothings posted at December 16, 2004 12:09 AM

>"Tasty 2 x Biscuit + Darren +", which doesn't look quite as outlandish.

It's easy to make outlandish:

Darren Biscuit 2 Tasty x + +

(which is just what Eric had with the data reversed)

Anyway, I think the "point" of this strip isn't actually all that significant. It simply reflects the huge difference between a fixed, read-once article and a frequently-updated, frequently-read article. It's inconvenient to have to go forwards through the whole thing every time, so we start at the end and go backwards. If there was a good automated way to make things go forwards but allow us to continue where we left off, we might actually still leave them in forwards order.

(Lord knows I actually read things like LiveJournal by going back through everything to the first unread, then going bottom-to-top through the entries so I read them in the right order. Slashdot is the only thing I'm willing to read backwards since it's rare that stories reference other recent stories.)

I think the actual amazing screwup of blogs isn't reverse-chronological order posting, it's comment pages that put the author of the comment at the bottom of the comment, instead of the top, which happens on about 95% of blogs. So you read "I" in the first paragraph of a comment and have no idea who "I" is. Is this Eric following up at last? Or just some random yahoo? Of the 8 comments existing as I type this, 7 of them have "I" or "me" in the first paragraph, 6 in the first sentence.

Comment from: Tragic Lad posted at December 16, 2004 1:54 AM

I've been reading a book by Tom Peters, Project '04. The book is supposedly written in a 'hyperlink' style, the author stating that he'll return to writing linear books when the world returns to a linear fashion. But you, know, I found it to be clumsy and disorganized as hell. There's some nice nuggets in the book but without focus or direction of 'cause then effect' they're lost in the hodgepodge.

I don't think the traditional layout of a book will ever seem screwy, as that layout and linear fashion to delivering a message is the optimal use of that technology.

Having the most recent entry at the top of a page is the most optimal use of blogs - though I don't see why there can't be a means to letting us, the readers, choose our preferred means of reading entries. It's all in one big database - why can't we choose how we prefer it to be spit out.

Comment from: Ray Radlein posted at December 16, 2004 3:51 AM

I'm almost afraid to find out if the Royal Canadian Artillery is still using the HP-41 for gun calculations.

Well, for the longest time, the fourth backup navigational computer on the space shuttle was an HP-41C hooked up to an IO loop and running some program that was, in all likelihood, knocked out by some guy at JPL in his spare time, or some such.

I'm almost certain that the shuttle never experienced a failure of all of the systems in front of it, and I suspect that it was replaced when they upgraded the shuttle fleet a few years back; making it yet another of the myriad NASA systems to have been designed, deployed, retired, and replaced without ever having been actually used.

Comment from: empyress posted at December 16, 2004 8:54 AM

HP48GX = my friend in engineering. They do still use RPN, so the first week is spent getting used to that. Although that is no longer the current model. And I have, admittedly, been out of school more than a year or two!

I agree with Tragic Lad - any books I've read that use "links" (okay, I've only read one, but it's not the one he mentioned, and I forget the title) are unweildy at best. The traditional "book" format is not going away any time soon.

But it was a great comic - emphasising how out of touch with the "offline" world some people get.

Comment from: Darth Paradox posted at December 16, 2004 9:47 PM

> any books I've read that use "links" (okay, I've only read one, but it's not the one he mentioned, and I forget the title) are unweildy at best.

In terms of fiction that is intended to be read in its entirety, this is certainly true. However, any manner of reference book is certain to need "links", and book format may in fact not be the optimal format for that. I remember - back when I used paper encyclopedias at all - taking down one volume from its rather high shelf, looking up the entry I thought I needed, discovering I actually needed a different entry, going back and getting another book, etc. Hyperlink technology is a very effective means for that sort of thing.

Another instance I've seen - and one I particularly enjoyed - was my family's geneaology book (published in 1932). Each entry had a unique identification number, and each entry would refer to the person's family by name and number, making it easy to do cross-references. I have to imagine that at some point, all this stuff was laid out on index cards, numbered and sorted, etc.

But fiction... we read things linearly, and as we experience things linearly, we don't really necessarily have the mental structure to be able to read a "linked" book like that (if I understand the way it's laid out, having not read it) and comprehend easily the overall structure. In any case, it takes work. And expressing multilinear, crossing story elements in a singly linear fashion is one of the things that a good writer ought to know how to do. Rearranging the order of things in a story is a long tradition, dating back long enough that the Greeks had a word for it, and can add points of suspense or irony to a narrative. (Consider the sort of story - I suspect we've all read something like this, but I can't really come up with an example - where we know the resolution of a story before knowing the details, and a flashback to a key part creates a sense of dramatic irony as the characters undertake an effort that we already know to have [succeeded|failed]. The aforementioned book (Project '04?) strikes me (in a very unfair review, having not read the book or even laid hands on it) as either a gimmicky attempt to lay out a story differently, or just a copout to avoid the challenge of writing a multi-threaded story in a single narrative.

But then, I'm no writer. *shrug*

Post a comment

Thanks for signing in, . Now you can comment. (sign out)

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Remember me?