Hacker News new | past | comments | ask | show | jobs | submit login
An algorithm for the Names at the 9/11 Memorial (newyorker.com)
154 points by erehweb on May 9, 2011 | hide | past | web | favorite | 47 comments

"Even so, the first few computer scientists and statisticians the foundation got in touch with said that it couldn’t be done."

Really? Wtf kind of incompetents were arguing that it couldn't be done?

I'm not arguing that it was or should have been easy, of course. It sounds like a fascinating challenge. And I'm also not claiming that any computer scientist, even a good one, could have solved it. But any vaguely competent computer scientist should have been able to say, "that's a cool problem, and you need to talk to a graph drawing expert."

Of course, this could have been a case of the source saying "they said it was hard" and the writer mangling that into "they said it couldn't be done". But still.

Of course, this could have been a case of the source saying "they said it was hard" and the writer mangling that into "they said it couldn't be done". But still.

I think it is more likely they said "It isn't a well-defined problem."

It's possible that the problem was originally stated in such a way that there was no solution. Perhaps the original statement of the problem was that each name must be adjacent to every name in its cluster, and readable. But the cluster of 658 people at Cantor Fitzgerald probably makes that impossible - how do you have 658 names all readable and adjacent to each other (in two dimensions)?

Perhaps they really answered "as the problem is stated, it can't be done, without compromises."

Then again, I tend to give computer scientists and statisticians the benefit of the doubt (and not so much the writer).

If I had been asked, my reply would have been "It can't be done. If you want help from me or someone like me, you'll need to make a loose ranking of the qualities of the ideal solution, and we can try some approximations".

I might even have left the latter sentence out, since it's clearly implicit. Maybe whoever they found left it out or maybe they didn't even consider it, but the artist could have made the addition too. I suspect the first few people just weren't interested.

"fascinating challenge" != "publishable"

It's possible that any academic asked might have concluded that there wasn't a general publishable result to be had from this work and therefore that it wasn't worth doing. Just because something is interesting doesn't necessarily mean it is of any interest from a research perspective.

[NB I left academic research in CS for pretty much these reasons - I like interesting stuff].

There are plenty of people that know computer science that will do contract work.

From the sound of it, the original request was probably stated as something massive and NP hard, he asked for the 'best' placement and was told it would take more time than seconds since the big bang to locate the best placement.

Then somebody actually talked to him and took the time to realize he didn't need the optimal solution.

Actually, solving most practical instances of NP hard problems is doable and quite tractable.

Obviously it depends a whole lot on the scale. They gave some idea of number of constraints required to be satisfied, but it's not clear to me what the actual constraints were. I think it's not outrageous that on scan it might have sounded intractable.

Actually, solving most practical instances of NP hard problems is doable and quite tractable.

Isn't that a tautology? Because if it wasn't doable then it wouldn't be practical.

Practical here means: Occurs in practice. And that they are optimally solvable in practice is not a tautology. It could be that you would need human insight (like you still need for most proving interesting mathematical theorems), or that people would only solve approximations. (People do solve approximations, too.)

As someone who has had research written about in popular media, I would bet an enormous amount of money at very unfavorable odds that the words "it couldn't be done" were not uttered by the CS researchers but instead are the voice of the New Yorker author.

I wonder if the coder/s would open source what they did. I imagine its purpose is unique enough that releasing the code wouldn't cause any "now we won't potential clients won't need to hire us" issues, and it could be quite interesting.

Edit: Any chance of explanation for the down voting to -1? I mean, I assume it's for thinking my question moronic, but would like to actually hear why that might be...

When you get downvoted, just wait a bit. It seems in the last two years or so we acquired some trigger-happy idiots. It still usually balances out in the end.

I don't really have a problem being downvoted, and I've certainly seen what you describe happen many times in the past. I was asking not so much to prevent downvoting, but because I couldn't see what I'd said that could lead people to want to downvote, and hoped to have a chance either to win them over or have them win me over.

Courtesy of the Metafilter thread (http://www.metafilter.com/103302/meaningful-adjacencies), a longer article:


And a website to explore the names and some of the stories and links between them:


What a fine combination of typographical aesthetics and algorithmic computer science! Makes me proud to be associated (however distantly) with those professions.

As for finding peoples' names, there's a strong symbolism it making it hard. To search for a name is to honor the efforts of the recovery workers who searched for remains.

At any rate, let's hope there's no more need for memorials like this one.

After all this, so-called aesthetic considerations make it impossible to find a name you are looking for.

At the Vietnam war memorial on the national mall in DC, the names are listed by date of death, but there are phone-book style guides that list the names alphabetically with the wall location.

Presumably this will offer something similar.

In fact, I wouldn't be surprised if it was just touch screens showing the names.911memorial.org page. What more do you need?

I'd bet there will be kiosks on site to find names, and there already is a website to find names that will point you to the panel: http://names.911memorial.org/

No, not impossible, just difficult. Because there are so many people being memorialized. I believe this is intentional.

Is the source/a paper/a blog post going to be released so that we can look at (read: poke holes in) it?

It's not like there's a "right" or "wrong" answer. It's more art than code, and what really matters is the final result.

i completely agree. i didn't mean that it would in any way be wrong, i am just really interested in how they did it, and wanted to look at the source.

yet another example of a need for a sarcasm punctuation mark.

Why would you add the parenthetical statement "read: poke holes in" if you were "really interested in how they did it"? Not only does it not really make sense (how do you "poke holes" in art?), it makes you sound like a dick.

Why is poking holes automatically negative?

Since it's to honor the lives of the victims, why not just sort them by date of birth?

This is way off topic, but if you use a diaeresis in "coöperation," you are probably an asshole. Stupid New Yorker.

I was going to comment exactly this.

I can't believe people are (ostensibly) paid well for this kind of writing. It meanders around and takes the longest path possible to what is essentially a paragraph or two of actual information.

Sometimes, writing is about more than just "actual information."

If this kind of writing bothers you, you definitely don't want to subscribe to the New Yorker.

I agree that there is some allowance for creative expression, but go back and re-read the first paragraph of that article. Does it add anything at all to the story?

In my mind it doesn't. With so many other things in life vying for my time I'm loathe to read and digest such fluff. It reminds me of what I used to do in school to boost my word count on papers. I'm sure professional writers labor under similar circumstances at times.

Do they routinely add those goofy umlauts, too? In the English I was taught in school, we don't have such a character.

Each requires the coöperation... even others coördinating a response

Perhaps I'm over-sensitive, but this seems rather elitist.

Now, combine this with the citation of "A same-sex couple and their three-year-old son..., certainly, belonged together." Why mention their genders? Either the writer has an unhealthy fixation on other people's sexual preferences, or, worse, thinks that the fact that they're gay entitles them to special consideration. Not good either way.

I don't think I like their style.


Contrary to popular belief, the English language has no official single governing body. Either are considered to be "correct".

Contrary to popular belief, the English language has no official single governing body.

Nor did I claim that it did; linguistics is a descriptive science, not a prescriptive one. It describes the way people use the language. I've never seen English with umlauts, outside of a few heavy metal bands [1]. The link you provided would seem to agree with me, showing the non-accented spelling outnumbering this one by some 500:1.

[1] https://secure.wikimedia.org/wikipedia/en/wiki/Metal_umlaut

"I've never seen English with umlauts"

Nobody cares.

yes, the new yorker always uses tremata/diaereses where most people would use hyphens (if anything).

I thought the "meandering" was actually a lucid explanation of the difficulty in choosing a suitable order of names in what is a highly sensitive location.

Anyway, I thought you chose an overly verbose way of saying "tl;dr".

To clarify on what others have replied to you about it being how The New Yorker does it: a piece like this is a feature, not news. Nobody reads the magazine (or their website) to get the quick facts, they read it because of the high quality writing and editorialising.

Your mistake is thinking the article is about the computer algorithm. It's about the journey.

I was just disappointed it didn't say much about the algorithm. It read like the intro to a longer piece.

Don't downvote people because you disagree with them, dagnammit.

His is a valid position, he explains his reason for holding it even. Please disagree, but recognise that his view is worthwhile and well presented given the context.

His position may be valid, but that doesn't mean it adds anything meaningful to the discussion.

I thought the point about creative writing being more than mere presentation of information was a worthwhile one to have brought out.

it's the new yorker. it's for reading on the john (on an educated, literary, liberal, and possibly jewish, john, i guess).

and i would have loved that job. if anyone has anything similar, contact me :o) (i can't believe people said "it can't be done")

Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact