0979: "Wisdom of the Ancients"

[phillipsjk]

melts from the shadows

Don't know if it is relevant, but Google only indexes terms it expects users to search for. If your query looks "strange", but is close to another query they expect, they auto-correct it with no option of using your original query unless they expect the "misspelling" and have it indexed.

I found this out while trying in vain to search for the Schlock Mercenary strip explaining why trademarks would be respected even in the face of replicator technology (neither "trademark" nor "replicator" appear in the transcript). In frustration, I tried searching for a term that I knew for a fact was on the schlockmercenary.com domain: "ominous hummm". That got auto-corrected to: "ominous hmmmm". The next day, when Google added "ominous hummm" to the index, I thought I might be going crazy.

The important take-away to understand is that Google is biased. It is biased to find new, popular information. I suspect the non-indexing of "strange" terms may be intended to discourage Google Hacking. However, the Bing practice of copying Google search results is annoying is edge cases such as those discussed above. If you want to be able to give people better search results, you can't just copy from your major competitor.

melts back into the shadows

[Uzh]

To me it sometimes looks like my specific fora have a mixture of http://xkcd.com/979 and http://xkcd.com/978. Very often I find a forum with the problem I've got and a reply: "Oh, visit this link" and it turns out that this link only refers to the original forum after some posts. Or sometimes they quote each other with only one solution which doesn't work for me.

I also encounter being linked to some strange search engines by google. In the referring text exactly my phrase is written bold, so I hopefully click the link just to see that they would do with every phrase...

Oh, I never thought I would do this, but GOOMHR!

Georg

[adaviel]

"Me, too"

I do try to post the actual solution. Usually what happens is I can't find a solution in the threads, fire off a question or two, and keep working on the problem myself and searching more industriously. Then find a solution, often before I've had any sensible answers. Or it may take so long I've forgotten where I posted. Last one was WebDAV in Windows 7.

Funny, I've never heard of stackoverflow.com in many years of searching problems online

[adaviel]

And another thing...

Recently I've been finding search engines hits for questions but no answers. People have been polluting the search space with ad-laden nothing. It's not even useless solutions, it's just null. HATE...

[brenok]

http://www.mirrorsoferis.com/forum/thread05232003a.html

"Life imitates Art far more than Art imitates Life"

The website recreated DenverCoder9 based on the comic;

[Uzh]

Recently I've been finding search engines hits for questions but no answers. People have been polluting the search space with ad-laden nothing. It's not even useless solutions, it's just null. HATE...

That's what I meant by my 2nd paragraph...

Never mind, fixed it.

Georg

[Ehsanit]

Just to say, there is a rationale behind the "Don't Bump ancient threads" rule. As time goes on, both the original poster and the helpers will tend to get bored and move on, possibly leaving the forum completely.
Most late posts are of the form "I too have this problem" with or without extra information. However when you turn up and post at the end, there's no-one there with particular expertise to help you. Meanwhile potential current helpers (who'd join a new thread) see a thread that already has a dozen replies, and assume that it is or will be sorted. You end up not being helped by anyone. A much better (moderator-proof) response would be to start a new thread, link to the old thread, and say "I already tried X, Y, and Z as suggested in that thread and the only result was A."

Of course this doesn't apply to those few noble souls who want to post the solution to an old thread, if not for the original querent at least for the benefit of posterity.

[bmonk]

But the most important measure has always been density and depth of imagery, as well as evocation of emotion.

I take exception to this. This assumes that poetry is a consistent form throughout history and across cultures. You simply can't make this case. Either poetry is a formal process, in which case you need formal terms to define it, or it is a cultural process, in which case you need to define it in terms of what it does for the people who make it and the people who hear it - or read it, but to think of poetry as something you read already takes you out of the older forms that you're trying to use to justify your position.
Either way, you can't justify that position. If you take the formal definition, of course, "density and depth of imagery" are nonsense, they have nothing to do with the matter at hand. If you take a cultural definition, you have a problem in that the ancient forms you're calling on to justify your position have nothing to do with what poets did two hundred years ago, or with what they do today.
I've given you an easy out, of course, which is to simply say "for me, poetry is X" and be done with it, but if you're going to insist that there's a correct definition here that suits all parties in the discussion, I'm afraid I'm going to make you work a little harder than that.

There is at least one other possibility:
Perhaps "poetry" is a fuzzy concept: there is no one definition that captures it precisely, and no two people will agree on it exactly, but most people will have a broad concept of the whole.


Oh, yes: searches for things composed of popular words are frustrating. It gets even worse when the site search engine removes words that are too popular, and then refuses to search for you at all unless you supply at least one word.

[chris857]

Only on xkcd can you start a thread about code forums, only to discuss the finer points of poetry.

[tubegeek]

One time I was on a woodworking forum and a poster had a question about a type of stain that he was having difficulty with. The chemist who developed the stain answered his question, within a few hours.

That's my favorite example of the opposite of this comic.

[jpk]

But the most important measure has always been density and depth of imagery, as well as evocation of emotion.

I take exception to this. This assumes that poetry is a consistent form throughout history and across cultures. You simply can't make this case. Either poetry is a formal process, in which case you need formal terms to define it, or it is a cultural process, in which case you need to define it in terms of what it does for the people who make it and the people who hear it - or read it, but to think of poetry as something you read already takes you out of the older forms that you're trying to use to justify your position.
Either way, you can't justify that position. If you take the formal definition, of course, "density and depth of imagery" are nonsense, they have nothing to do with the matter at hand. If you take a cultural definition, you have a problem in that the ancient forms you're calling on to justify your position have nothing to do with what poets did two hundred years ago, or with what they do today.
I've given you an easy out, of course, which is to simply say "for me, poetry is X" and be done with it, but if you're going to insist that there's a correct definition here that suits all parties in the discussion, I'm afraid I'm going to make you work a little harder than that.

There is at least one other possibility:
Perhaps "poetry" is a fuzzy concept: there is no one definition that captures it precisely, and no two people will agree on it exactly, but most people will have a broad concept of the whole..

Yes.

I've given you an easy out, of course, which is to simply say "for me, poetry is X" and be done with it, but if you're going to insist that there's a correct definition here that suits all parties in the discussion, I'm afraid I'm going to make you work a little harder than that.

I'm glad you took the easy out, it's actually the sensible solution in this case.

[Dampening]

And the saddest part is, of course, finding that your problem has existed for a full eight years before you encountered it.

[Deadcode]

Don't know if it is relevant, but Google only indexes terms it expects users to search for. If your query looks "strange", but is close to another query they expect, they auto-correct it with no option of using your original query unless they expect the "misspelling" and have it indexed.

Actually Google lets you use "strange" queries without any auto-correction. Just enclose your search terms in quotes, and Google will search for exactly what you typed. (You used to be able to do this by inserting a "+" in front of a word, but that has recently been disabled. As far as I can tell, enclosing in quotes does everything a prefixed "+" used to do.)

Another thing enclosing in quotes does (and a prefixed "+" used to do) is force Google to return only web pages that actually have the search term in the contents of the page. Without the quotes, it often returns some results that have the search term only in pages that link to them.

[Euler]

There's one thing worse: a single followup reading just

Google it.

Or the infamous "Let Me Google That For You" (http://lmgtfy.com/?q=solve+problem+%22x%22)

I hate the "Solved the problem, thanks!" and nothing on how they solved it...

Even worse: http://bit.ly/tRS44R

[Jorpho]

Don't know if it is relevant, but Google only indexes terms it expects users to search for. If your query looks "strange", but is close to another query they expect, they auto-correct it with no option of using your original query unless they expect the "misspelling" and have it indexed.

Actually Google lets you use "strange" queries without any auto-correction. Just enclose your search terms in quotes, and Google will search for exactly what you typed. (You used to be able to do this by inserting a "+" in front of a word, but that has recently been disabled. As far as I can tell, enclosing in quotes does everything a prefixed "+" used to do.)

Another thing enclosing in quotes does (and a prefixed "+" used to do) is force Google to return only web pages that actually have the search term in the contents of the page. Without the quotes, it often returns some results that have the search term only in pages that link to them.

They've also added a new option called "Verbatim" in the Search tools in the lower-left that lets you accomplish the same thing without having to put quotes everywhere.

[phillipsjk]

Don't know if it is relevant, but Google only indexes terms it expects users to search for. If your query looks "strange", but is close to another query they expect, they auto-correct it with no option of using your original query unless they expect the "misspelling" and have it indexed.

Actually Google lets you use "strange" queries without any auto-correction. Just enclose your search terms in quotes, and Google will search for exactly what you typed. (You used to be able to do this by inserting a "+" in front of a word, but that has recently been disabled. As far as I can tell, enclosing in quotes does everything a prefixed "+" used to do.)

I am fairly certain I did use quotes. It is more likely that I was actually using Altavista, which is Yahoo!, which is Bing at the time.

Normally, you are given the option to search with the mispelled term. In the case I describe above, there was no such option.

[teelo]

I feel like the left side of the comic desperately needs to be a poem.

Perhaps something like...

I hit google to find the answer to my error
But there was only one result to my terror
Never have I thought that we were so alone
As when I found the thread it was your own
To solve the problem I look unto thee
But no response to your thread since two thousand and three

[Eugo]

And the saddest part is, of course, finding that your problem has existed for a full eight years before you encountered it.

Accidentally, for eight days now I have a tab open on a forum of the manufacturer of an ActiveX component, of a thread started in 2006, amended in 2007 and now by me. The bug (which seems to not be qualified as such by the manufacturer) is still there. But they have another, nastier bug that survived from version 4 into version 16 (and probably 17, coming out soon without a hint that they have even considered that... unnecessary feature that can't be turned off). On that one I have pretty much given up. IOW, I not only had a GOOMHR, I had an ongoing situation (actually two) of the kind - would this qualify as "stop peeking into my browser's tabs, Randall"?

On the BTDT front, my best experience was that I had a problem that nobody else had. Just one guy with the same problem (me, seven years ago), but luckily someone had a solution and posted it in the same thread (me again). The deja vu was just mild, and actually hit me only a day later. (anyone else with this delayed deja vu syndrome out there?)

[DisreputableDog]

However has anyone else besides me tried to email or message the person, find the account is dead, then use google and Internet Archive and other trickses to try to track down the current email address or website of anyone using the same handle, and ask them if they remember how they resolved that question years ago? One time I actually did manage to get in touch with the person, although they had only the vaguest recollection of the issue.

How about when you're looking for a foreign music tape/possible cd that was your childhood favorite.
So you start your Google search in that country, but on the way discover a number of old news articles mentioning that he was suspected of murder.
And you figure, "So he's probably this -other- guy with the same name, now several thousand miles away, cause how do you live in the same town with -that-?"
So you trip over his Youtube user name and watch the videos, containing family pictures (because you know what he looks like at a young age)....
and you just image him several years older and flip between pages back and forth to confirm it's him.
And what do you know, his Youtube account has a comment by someone who looks like a friend, with a Facebook account.
And yes, yes he has a Facebook account too! So you send him a polite "Apologies that this is so strange but I'm looking for this cd...." message.

It's been a year. No, he hasn't written back. *Sighs* Back to square one.

[dchorror]

Here's one that I didn't see listed but has happened to me a couple of times.


Asked a question
Someone gave me a link to the solution
Which was just another link to the solution
Which was really just another link to the solution
Which was an article ending in a link to the solution after complaining about it
Which was finally a link to a download


For Spyware.

[Fysicist09]

This may be the strongest GOOMHR moment I have ever had.

"Wisdom of the Ancients" is just the perfect title, and it evokes such a sense of poring over some long-forgotten information to find the exact nugget that you need to solve the problem - but the ancient writer was stopped by some malicious force. (The Castle Aaauuuuuuggggghhhhh, anyone?)

Oh scholars of Yore, why have you forsaken me?


Edit: Holy crap, there's a DenverCoder9 on StackOverflow. At first I assumed it was another back-dated tribute like the mirrorsoferis link that brenok metioned above (at http://www.mirrorsoferis.com/forum/thread05232003a.html), but upon further inspection this StackOverflow user has a history of posting 15 questions and 51 answers over the course of almost two years - some of which have responses from other users. Check it out at http://stackoverflow.com/users/297981/denvercoder9

For my money, that seems a little too elaborate to be a reaction to this comic. HAVE WE FOUND THE *REAL* DenverCoder9?

[Sgore]

Back in 2005, I posted a tech issue I had with an already ancient computer game on another message board.
The thread got two replies, one of which was mine, and died little over a week later.

Cut to February of this year. I get curious about this issue again and go searching for help with it. I find my way back to my original post by sheer coincidence, and someone has posted a solution in it that exact day, over 5 years later. Thankfully, I also happen to still be an active member of that message board community, so I was already logged in when I found it, and able to post a thank you quickly.

The whole single page exchange (involving unlocking an expansion pack/sequel for an old video game whose company went out of business) is right here:
http://www.gamingsteve.com/blab/index.p ... 7#msg28377

Just goes to show these stories sometimes do have happy endings...though they make take a while longer than one expects.

[Eternal Density]

This happens to me on occasion. But it doesn't make me feel 'close' to anyone.

[project2051]

There's one thing worse: a single followup reading just

Never mind, I figured it out.

"Never mind, I figured it out. The problem was in the uuggggggh"

[meerta]

The left side of the comic already is a poem. Poems don't need to rhyme or have strict metre (see Blake, Browning and so on)

This comic speaks to anyone who know to search for problems they're having. This thread alone has made me laugh. It's practically a manifesto.

[bmonk]

There's one thing worse: a single followup reading just

Never mind, I figured it out.

"Never mind, I figured it out. The problem was in the uuggggggh"

Isn't that supposed to be Arrrggghhh??

[Skydiver]

The point is to have a very simple system for collecting problems and just their solutions in one place for easy access.

What do folks think of the idea?

Such a site would be a valuable resource. But.

What would motivate people to post there? I'm afraid it would end up being a very empty site. On sites such as expertsexchange, the person asking the question is obviously motivated by getting a solution to his problem. The people answering the questions could be motivated by:
- being a part of the community
- getting more points/respect/whatever metric they use on the site to measure the "best" answerers
- genuine desire to help the person with the problem.

On ProblemSolved, none of the above would apply. The third one partially, yes, but posting an answer in the hopes that The People Of The Future find it valuable is nowhere as powerful a motivator as a person having the problem right now.

So, I'll just throw in some random thoughts about solving the above problem:

I guess you could add a voting mechanism, something like "Did you find this solution helpful", but I don't think that's enough.

Maybe you could add a possibility for the users of the site to discuss the solution? Or a possibility to somehow tag the solution as being related to a new problem... on ProblemSolved's sister site SolveMyProblem? And then, if someone's problem on SolveMyProblem gets solved, there would be an option to add the solution to ProblemSolved?

[Kailen]

This comic hits far, far closer to home than I'd like it to.

[scottyb]

As of 11:12PM US CDST a Google search for "DenverCoder9" returns no results.

As of 4:00 AM GMT -3 November 18 of 2011, a Google search for "DenverCoder9" returns YOU.

Oh great. Now there are a million hits but they're all just references to the original, unsolved thread!

[scarletmanuka]

It's even worse when you're running localized software and all the error messages are in Norwegian. How the fsck are you supposed to google that and find a decent answer?

Google Translate?

- getting hits to that silly "experts-exchange" site, which is not helpful anyway;

Used to be you could see the answer if you scrolled down for about half an hour. They've tightened it up a bit now.

- whenever searching for something database related, I'm always getting hits to the many sites of that Oracle consultant with his face on every page.

Ah, Dan Burleson, how we love to see you repost the official description of every Oracle error, with no additional information at all.

I hear you, bro. 90 % of the times I need to ask a question in a forum, I get completely unrelated and useless answers. It seems to be too hard for most people to either answer the damn question or keep their fingers off the keyboard. And fair enough that people may not understand completely what I mean by what I say and thus ask me questions, but way too often they NEED to know why the Hell I need to do the things the way I describe.

IT DOES NOT MATTER, JUST ANSWER THE F***ING QUESTION!

If you read Raymond Chen's blog, you'll see that sometimes they have to ask why you want to do things that way because the original question doesn't make sense or isn't what you actually need, and many times they have to ask because
the questioner really is doing something bad. (My favourite example is here.)

How do you know it doesn't matter, if you don't know what the answer is? Conversely, if you know you're doing things in an unusual way and you have reasons for doing so, how hard is it to explain that? "I know you would normally do this via <X>, but that's not an option for me because <Y>" can go a long way. Of course there are always some users who'll just say "You should use <X>", but there's nothing you can do about that. That's all they would have told you no matter what you said.

[s271]

Real #denvercoder9 ?
http://stackoverflow.com/users/297981/denvercoder9
member for 1 year, 8 months
Or just nick changed?

[Hafting]

BTDT.

It's even worse when you're running localized software and all the error messages are in Norwegian. How the fsck are you supposed to google that and find a decent answer?

Luckily, in Gentoo I can always disable nls in the compiler options

This is easy if the error is reproduceable. (Or you wouldn't need to fix it anyway).

LANG=C command-that-fails-with-an-error-msg

This way, you get the error message in the program's original language, which almost always is English. No need to disable localization or disable your own language.

On the other hand, google works for all languages. I have often enough found answers in Norwegian.

[Hafting]

Don't know if it is relevant, but Google only indexes terms it expects users to search for. If your query looks "strange", but is close to another query they expect, they auto-correct it with no option of using your original query unless they expect the "misspelling" and have it indexed.

Actually Google lets you use "strange" queries without any auto-correction. Just enclose your search terms in quotes, and Google will search for exactly what you typed. (You used to be able to do this by inserting a "+" in front of a word, but that has recently been disabled. As far as I can tell, enclosing in quotes does everything a prefixed "+" used to do.)

Another thing enclosing in quotes does (and a prefixed "+" used to do) is force Google to return only web pages that actually have the search term in the contents of the page. Without the quotes, it often returns some results that have the search term only in pages that link to them.

Doesn't work - never did. try googling "int n=1;". You get plenty of hits without the semicolon. Now, this example was made up. But it is a hopeless situation, when you search for something obscure that almost matches something 50 000 times more common. The internet may have an answer, but google drowns it in obviously wrong hits they mistakenly think is similiar. Or try "beiber" - all you get is a bad singer with a similair name. Quotes won't help

[monotone]

^ I can't reproduce that probem. Searching "beiber" produces a link to 'Did you mean 'bieber'' then thecorrect results for "beiber".
Granted, many are misspellings of an irritating singer, but you can hardly blame Google for others' bad spelling. Your first example is because Google ignores punctuation to assist in the difference between say "Bridget Jones Diary" and "Bridget Jones' Diary". You can show this easily by searching for "!!!" - there must be hundreds of web pages with that in them, heck there's even a band called !!!, but Google returns no hits.

[RobFreundlich]

The point is to have a very simple system for collecting problems and just their solutions in one place for easy access.

What do folks think of the idea?

Such a site would be a valuable resource. But.

What would motivate people to post there?
...
- being a part of the community
- getting more points/respect/whatever metric they use on the site to measure the "best" answerers
- genuine desire to help the person with the problem.

On ProblemSolved, none of the above would apply. The third one partially, yes, but posting an answer in the hopes that The People Of The Future find it valuable is nowhere as powerful a motivator as a person having the problem right now.

I'd like to be able to say you're wrong and that people are inherently good and would be motivated enough by a genuine desire to help, but unfortunately, I suspect that's not the case. At least not for most people.

Personally, I like helping people and that would motivate me. In addition, if there were such a site, I'd post solutions to it in the hope of keeping it alive in case I ever needed to find a solution on it.

So maybe selfishness (of the form "hey, if I post here, maybe someone else will post something I'll need someday") would be the motivation for most people?

So, I'll just throw in some random thoughts about solving the above problem:

I guess you could add a voting mechanism, something like "Did you find this solution helpful", but I don't think that's enough.

People seem to be motivated by the (IMHO) silly badges and "membership levels" that other fora (and review sites and so on) have. In fact, there's a bit of similarity between ProblemSolved and (for example) the reviews people post on Amazon. IN both cases, they're someone taking time to tell the world about something, with no real payback. So maybe user ratings and badges and reputation points and other stuff like that would be useful.

Maybe you could add a possibility for the users of the site to discuss the solution? Or a possibility to somehow tag the solution as being related to a new problem... on ProblemSolved's sister site SolveMyProblem? And then, if someone's problem on SolveMyProblem gets solved, there would be an option to add the solution to ProblemSolved?

I like those ideas a lot. You could also provide a social-networking-like icon to make it easy for users on other fora (such as StackOverflow) to connect their problem with a ProblemSolved solution (instead of "Like this", call it "ISolved this").

Now I'll add a problem without a solution: I have very little time to work on this, and even less in the way of resources such as a hosting site (or $$$ to pay for one). So maybe the very first ProblemSolved solution should be "how to set up a ProblemSolved" site

[Euphonium]

Moderator: DON'T BUMP OLD THREADS MAKE A NEW ONE *LOCKS THREAD*

I have yet to see a valid explanation of this rule.

I suspect it's just for the moderators' own egos, so they can say "I moderate an Internet forum with 10,000 discussion topics" rather than "I moderate an Internet forum with 2,000 discussion topics."

[JimmyKills]

Two wonderful stories related to this from my past two weeks.

1. Several days ago my co-worker asked me for help with an error he could just not fix. After trying a few solutions he didn't think of we relented and Googled it. The only result was a forum post from 2002 with the same error. There was a single reply that made my co almost break down in tears "just google it." Made from his account there from back then.

Some sort of divine justice I suppose.


2. About three weeks ago now, I received an email which had been bounced through several old emails I had set to forward (4-5 total) with a post on an old forum I thought was obsolete. Someone had PMed me with a thank you directed at the solution my 16 yr old self had posted saying it was one of only a few hits on any search and the only one with an answer.

[SecondTalon]

1. Several days ago my co-worker asked me for help with an error he could just not fix. After trying a few solutions he didn't think of we relented and Googled it. The only result was a forum post from 2002 with the same error. There was a single reply that made my co almost break down in tears "just google it." Made from his account there from back then.

Some sort of divine justice I suppose.

...

That's beautiful.

[Euler]

Back in 2005, I posted a tech issue I had with an already ancient computer game on another message board.
The thread got two replies, one of which was mine, and died little over a week later.

Cut to February of this year. I get curious about this issue again and go searching for help with it. I find my way back to my original post by sheer coincidence, and someone has posted a solution in it that exact day, over 5 years later. Thankfully, I also happen to still be an active member of that message board community, so I was already logged in when I found it, and able to post a thank you quickly.

The whole single page exchange (involving unlocking an expansion pack/sequel for an old video game whose company went out of business) is right here:
http://www.gamingsteve.com/blab/index.p ... 7#msg28377

Just goes to show these stories sometimes do have happy endings...though they make take a while longer than one expects.

My god.......
Feelsgoodman.
You just made my day.

[Euphonium]

There was a single reply that made my co almost break down in tears "just google it."

What makes these sorts of responses even worse is that there is literally zero reason to assume the person hasn't already googled it.

A Google search for an answer, if successful, is much faster than posting on a forum (especially if it's one on which you are not already registered) and waiting hours or days (or forever) for someone to respond. So there's no reason not to think the person asking already did exactly that.