« first day (2519 days earlier)      last day (1020 days later) » 
02:00 - 16:0016:00 - 00:00

2:16 AM
@BalarkaSen This paper taught me one thing: It seems 2 forms can be represented as curves with a clockwise or anticlockwise sense/orientation
 
@BalarkaSen set theory contains memes i.gyazo.com/3f4e0580d3c4fdb85e16e50b3c7d18eb.png
succ
 
you need axiom of choice to do that union I think...?
 
$\mathcal L{\llcorner} \xi$
oh no
 
Lie derivative of some swiggly thing?
 
I am trying to get that corner to be the correct size
@Secret restriction of a measure by a current
geometric measure theory has the strangest symbols
they couldn't be happy with the usual ones
I can't seem to find this symbol
 
2:33 AM
The $\llcorner$ is written as if it is a binary operator based on the size of that script
 
mhm
 
3:24 AM
I need that space on my SSD...
 
 
1 hour later…
4:39 AM
@JohnRennie there is a cloud of fruit flies in my kitchen! I don't know why!
 
Make fruit fly risotto. Very tasty.
If a bit crunchy.
 
@JohnRennie :/
 
4:56 AM
Good morning @JohnRennie
 
Morning
 
@JohnRennie I did the apple cider vinegar trick
I hope it works...
 
@JohnRennie I still don't get cleared with *&a
&a means address of a
* before &a means &a is a pointer to some integer
then which is that intteger
 
Suppose you write int a = 1;
 
ok
 
5:04 AM
This allocates 4 bytes of memory and stores the value 1 in those bytes. The compiler uses a as a label for those four bytes i.e. when you write a the compiler replaces it with the value stored in the 4 bytes.
 
yes
 
The & operator returns the address of the first of those four bytes
So &a returns a number that is the address where the value in a is stored.
 
ok
 
&a is a type called a pointer because it points to where in memory data is stored.
More specifically it has type int*
So you could write int* p = &a and creates a new variable called p that contains the address of the first byte of a
ok so far?
 
ok
 
5:09 AM
If you have a pointer p then *p returns the value pointed to by p. It takes the address stored in p, looks to see what data is stored at that address and returns that data.
So in this case *p == a
 
yeah
 
But p is just &a, so that means *(&a) == a
 
oh great
Thanks
 
You don't need the brackets because the compiler processes the & before the * anyway, so *(&a) and *&a are just different ways of writing the same thing.
 
ohk
got it
 
5:14 AM
Why do people insist on starting numbering at 0?
 
@0ßelö7 In computing, or generally?
Fair enough :) Interestingly, after space was measured flat, the entire Wikipedia was carefully edited out to remove references to the initial size of the universe. One can hardly find either "a point" or "infinite". The dilemma is that "a point" contradicts the consensus theory conclusion while "infinite" is singular by mass and distance and therefore has no physical meaning. Either way FLRW is doomed, but no one has a courage to admit it. The Milne (Walker's teacher) model was ruled out, because it was singular. The clock is ticking for FLRW too. You should always choose physics over math :) — Victor Storm 29 mins ago
Looks as if we have someone to hold the fort until John gets back
 
@JohnRennie computing
 
Because the lowest value of a byte is 0B00000000
 
That seems like a very bad reason
Humans start counting at 1
:40227075 Einstein and the evidence
 
@0ßelö7 noooooooooooooooooooooooooooooooooo :-)
 
5:22 AM
@JohnRennie looks like it's time for me to make a new friend. How do invite to the chat?
 
I don't think you can, apart from commenting to that thread and saying come join the chat
 
You're lucky I can't @ people in comments while on mobile
Pray that I forget about this tomorrow...
Gnight
 
5:51 AM
0
Q: Mainstream physics closed as non-mainstream

WillOThis (admittedly naive) question (asking whether the existence of tachyons would have observable consequences) seems to me to be clearly about mainstream physics, though it was closed as non-mainstream. If the answer is that the existence of tachyons would have observable consequences that contr...

 
6:46 AM
in this why we cannot use like void swap(int a,int b)
 
Remember that a and b are just labels for values. So suppose a=1 and b=2 then if you used swap(int,int) you would effectively be calling swap(1,2)
And that doesn't do anything useful
 
then how pointers are useful
in the int main() in place of a,b we have addresses
i.e &num1 and &num2
 
What you want to do is swap the contents of two locations in memory. That is take whatever is stored in the first memory location and swap it with whatever is stored in the second memory location.
 
yes
 
So the arguments to swap need to be locations in memory.
i.e. the arguments to swap need to be pointers
 
6:54 AM
if argumentns is pointers then inside swap fuction we should use addresses that is only a and b
to swap
 
Inside the swap function youwant to swap the contents of a and the contents of b
 
we would swap the addesses of a and b
 
Suppose you have a pointer int* pa then the contents of the memory pointed to by pa are given by *pa
 
yes
 
So the line t = *a takes the contents of the memory pointed to by a and stores that number in t
 
6:57 AM
@0ßelö7 I want to know Chapter 1
Not sure about you
 
Then *a = *b takes the conents of the memory pointed to by b and stores it in the memory pointed to by a.
 
so in main function a is ponting to num1
Am I correct ?
 
And finally *b = t takes the number that was in the memory location pointed to by a and stores it in te memory location pointed to by b
@Koolman correct
 
oh its interesting :)
 
I'm not sure if this helps, but I'll mention it anyway ...
It's conventional to write function declarations in the form used in that article i.e. swap(int *a,int *b) but I think it's better to write swap(int* a,int* b)
 
7:04 AM
void swap( int &num1, int &num2 )
{
int t;
t = &num1;
&num1 = &num2;
&num2= t;
 
This means exatcly the same, but it makes it clearer that the arguments are of type int* i.e. pointers to int.
 
if we are using swap function in main function then it would be like that
not the contents of the memory
 
I'm not sure what you are asking
You're suggesting we could define the function swap as you've written above?
 
@JohnRennie no
we define the function as void swap( int *a, int *b )
{
int t;
t = *a;
*a = *b;
*b = t;
}
but in main function *a is equivalent to &num1
 
If you're doing the swap in your main program you just use:
t = a;
a = b;
b = t;
 
7:10 AM
no first I am defining the swap function
in main function we have written void(&num1,&num2)
that should means &num1 is equivalent to *a
and &num2 is equivalent to *b
That is we can replace *a by &num1 and *b by &num2
 
@Koolman The main program contains swap(&num1, &num2) not swap(num1, num2)
 
Sorry I have changed it
 
@Koolman No, &num is equivalent to a not *a
Remember that in your function a has type int*
 
@JohnRennie Why
You mean &num1 is also a pointer in swap function of main fucntion
 
In your function a and b are type int* i.e. pointers to int
@Koolman yes. num1 is type int so &num is type pointer to int
 
7:17 AM
ok ,let me think more on it
got it
thanks a lot
16 mins ago, by John Rennie
It's conventional to write function declarations in the form used in that article i.e. swap(int *a,int *b) but I think it's better to write swap(int* a,int* b)
This is very useful statement
 
I always write my code that way because it makes it easier for me to understand
It makes it clear what the types of a and b are
 
Yeah , its a better way
 
7:41 AM
Does someone know about the flow charts of manchester series books?
Why are there numbers not being chapters of the book? and names of the numbers dont match chapter names?
 
Sid
@Koolman now. Try to swap without using a third variable
 
there is already a third variable t
 
Sid
Well, write another piece of code which swaps but without using a third variable
That is another useful program for beginners to know. I have heard Companies even ask about that in interviews
 
x=x+y;
y=x-y
x=x-y
 
Sid
:-). Yep
 
7:53 AM
Excellent. A saving of four bytes of memory at the cost of comprehensibility and with the added excitement of potential overflows :-)
 
But actually I want to learn about pointers
 
Sid
@JohnRennie is it possible to golf that even more? I don't think it is, but would like to know if it can be.
Of course, we can make it as ugly as we want to just save a few bytes
 
 
2 hours later…
Sid
9:32 AM
3
Q: Why, exactly, can't two field lines cross?

AnkoganitMy physics textbook says two electric field lines never intersect. Their explanation runs somewhat like this: If two field lines crossed, there would be two different directions to the electric field at the intersection point, which is impossible by definition. A similar explanation is prov...

How is that ^ a duplicate?
I think that's a valid question.
 
10:14 AM
@Sid Because...it's a duplicate? If you are not content with the given answers to a question, you're not supposed to ask the same question again, but offer a bounty with the "more detail needed" reason.
 
 
3 hours later…
12:46 PM
@ACuriousMind Uh, what exactly is the Calabi conjecture? The wiki page is not what Calabi claimed, apparently
 
Hello
I conjecture that a person named Calabi exists
 
@0ßelö7 If you know that what Calabi claimed is not what's on the Wiki page, then you know more than me.
 
Aw, I was going to conjecture that there exists a conjecture that Calabi came up with a conjecture :P
 
I am reading Aubin's proof and he says the Calabi conjecture is that for each form representing the first Chern class, there is a Kahler metric with that Ricci curvature
but Wiki says that it has to do with the Kahler form being a representative of the chern class, I think
oh no, I just can't read
 
@0ßelö7 Apart from a missing "unique", that appears to be what Wiki says
 
12:50 PM
@ACuriousMind Well no, they also have a condition that says the Kahler forms have to be cohomologous
@ACuriousMind Aubin has uniqueness implicitly
But the fact that the new Kahler form is cohomologous to the old one...?
 
Sid
1:16 PM
@ACuriousMind the guy does not understand the answer. And they also show that a common tangent is possible for two intersecting curves. Nowhere do I see anything that can help them understand where there is a fault in their argument
 
1:36 PM
Someone should post a rigorous mathematical answer to that question
 
@BalarkaSen uniqueness for ODEs
 
yup, picard-lindelof
 
Picardy lindyhopfer
@BalarkaSen the answers to the question are hand wavy
 
yeah all the answers are bad
bad bad bad
 
@BalarkaSen oh I'm gonna take a while to get to Jost. Need to finish my current project
 
1:40 PM
For sure
 
@JohnRennie the flies are gone. One drowned
I am scared I will find his brothers nesting in a potato or something
 
I posted a comment
 
Really? An arrow on the top?
 
lol
 
What is wrong with mathbf you communist
 
1:43 PM
physicists like arrows
 
You don't like thickness I see
 
i need to give them the impression that my comment is welcome to them
 
Did you get my succ meme
@BalarkaSen mathbf E is the correct symbol
 
I am angry and heartbroken that Yashas's answer is unfortunately completely wrong
Two curves can easily have the same tangent vector at a point
The issue is it can't have the same tangent vector everywhere locally without being the same
DISAPPOINTED
 
Did you leave a comment?
Indians can't into analysis kek
 
1:49 PM
i want to leave a comment but i feel like that'd be trolling too far
 
Is that even possible?
 
good question
 
@Sid That the answers to the other question are wrong as that question shows still doesn't make it a different question - I'm not sure what your point is
 
Sid
The OP is claiming it's a follow up question to the original question. And I don't disagree. They are asking to prove the original point while claiming the original reason saying about different direction of field at intersection point has a flaw
Pretty sure, that's a different question than the original question
 
I posted an answer on the comment section, fear not
 
2:03 PM
@Sid It's not a follow-up question. It's the exact same question, just with the added information that one particular kind of answer is not sufficient.
@BalarkaSen How about you post that as an actual answer to the original question, instead?
 
Can't, I don't have 10 reputation points
It's a protected question
Also I don't think physicists would like me writing a paragraph consisting entirely of a technical ODE lemma and say how that answers their question :P
 
Lmao
@BalarkaSen I have lots of answers like that
They usually have 0 points
 
@BalarkaSen Well...get 10 reputation points then :P
 
Sid
@ACuriousMind and that particular type of answer is what is accepted by the OP in the original question
 
@Sid Acceptance means nothing but "this helped the asker of this question the most".
In particular, you should not think of it as a "this is the most correct answer" or a "this is the most detailed answer".
There are a few questions where blatantly wrong answers are accepted, we can't do anything about it.
 
2:10 PM
@0ßelö7 Unsurprisingly
why are we even here
the physicists won't accept us
 
Sid
@ACuriousMind and that particular answer didn't help the OP of this question.
 
@Sid Yes, also true. How does that make the questions not duplicate?
 
@BalarkaSen most of the people in this chat are mathematicians
And the math chat is annoying
 
ah well true
 
Sid
I see your point
 
2:13 PM
@BalarkaSen If you drop a bunch of mathematical jargon in response to a non-jargon-laden physics question, the burden is on you to convince the reader how that jargon relates to the question ;P
 
@ACuriousMind because otherwise OP has to rely on an old question with only incorrect answers
I find the SE system of duplicates asinine, honestly
If the old question is dead and doesn't have satisfactory answers, why make the OP suffer?
 
Anonymous
@0ßelö7 I agree with that
 
Sid
In any case, the OP didn't benefit much from visiting this site. In which case, are we really helping people clear their confusion?
 
Anonymous
The duplicate system needs some changes
 
@0ßelö7 The goal is not to make anyone suffer. The goal is to collect all the answers in a single thread, and also to have a long-term archive. There is no such thing as a "dead" question
 
2:15 PM
@ACuriousMind Which is exactly why I do not want to post an answer. The comment was made in half-jest to see if I can get some physicists triggered.
 
@ACuriousMind Do doubts get bumped if they are marked as duplicates?
 
Anonymous
Marking questions as duplicates shouldn't prevent new answers. Moreover, hardly people will go back to old questions (to which the current question has been linked) to answer them. People prefer answering new questions.
 
Doubts should get bumped and have a note on them that some other person is looking for a better answer
 
@Sid This is a frequent point of confusion: The primary goal of SE sites is not to help individuals. The primary goal is to generate a database of knowledge that does not require the searcher to click through a dozen different threads, as it happens with forums.
 
@Blue yeah, even better: have answers on duplicate doubts go to both the original and the duplicate
Then once the OP has the answer they want they can close their doubt and everything merges
 
2:17 PM
@0ßelö7 Well, since the question marked as duplicate appears in the "active" list, it usually sends some traffic to the original. When I get upvotes on old posts it's often because the question was linked as a duplicate to a newer one.
 
Anonymous
@0ßelö7 Agreed. We should raise this on meta. Really
 
@Blue Well, but the SE system intends to get people to answer the old ones, too.
 
If it's a duplicate then it shouldn't matter which one is answered
Just merge them once an answer has been accepted
 
@0ßelö7 But then you have to click through all linked duplicates
 
I literally said merge them in the end
 
2:19 PM
Merging does horrible things to the database
In particular, what happens to the question post? Duplicates are usually phrased differently, if you "merge", which version of the question do you take? Who gets the rep if the merged post is upvoted later?
 
Just have the duplicate doubts as dropdowns and some marker on the answers saying which doubt they originally responded to
I mean cmon have some creativity
@ACuriousMind rep is my lowest priority here. Do whatever
 
Sid
Muricans have started using "doubt". What has the world come to?
 
You're free to propose new systems for duplication on Meta Stack Exchange, but I promise you they either already have been proposed and shot down or will then be shot down.
 
Oh I don't believe for a second my proposal would be considered
But it is a good one
 
For a different site, maybe, but not for what SE sites want to be.
 
Sid
2:25 PM
@ACuriousMind Okay. Did anyone try to clear the confusion of the OP(of my linked question) in the original question?
 
Well it's not the first time I disagree with a fundamental SE principle
I actually have no idea why I'm here
@ACuriousMind is there a way to delete an account but keep the favorite doubts list?
 
@Sid No, apparently not. If you think this is an important confusion to address, you should offer a bounty requesting answers that do.
 
Sid
@ACuriousMind Ah. So, basically we have to bribe people to get our questions answered. Great.
 
@0ßelö7 is this you
 
Nope
@Slereah Are you still sending those things?
 
2:30 PM
I am not
I am at school
 
Ok I am too
 
To learn my ABC
Just kidding I'm slacking and writing a thing on the Alcubierre metric
 
You should buy that book I sent you
Having Sanchez in paper form would be cool
 
Anonymous
@ACuriousMind Okay, but I don't know what is SE's argument against allowing answers to duplicate questions. All duplicated questions could just be linked to one another. People could then look through all the answers and decide which one suits them the best.
 
A fine metric
yeah, though not right now
I spent a lot of mullah on antique computers lately
I must curb my spending
Plus I bought that Visser book
Which I look forward to
 
2:32 PM
@Sid In this case, the proper course of action is to offer a bounty, yes - what you have is not a new question, it's just one that hasn't been answered by someone pedantic enough. If you want to call that "bribe", you can, but I don't think that's how you should think about it since "bribe" carries with it some connotation of illegimitacy, while the ability to offer bounties is baked into the system.
@Blue The main argument is that clicking through a dozen duplicates to see whether the answer you want is in any of them is not what casual searchers want - most of SE's traffic comes from people arriving via search engines and SE naturally wants them to have a good chance to find what they seek on the page they arrive on
@Slereah But...why?
Nostalgia?
 
Partly
Also it is fun to code on them
u work close to the machine
 
@Slereah I bought the friedlander book on wave equations in GR
The one that Moretti always talks about
Gonna learn about the Hadamard form I hope
 
Sid
@ACuriousMind Okay. So, my question didn't have an answer in the original question. So, my question asking specifically what my concern is, is closed. I offer a bounty on the original question hoping that someone answers it while staking my entire reputation on the line.
And if no one answers it to my satisfaction, the rep is gone and I still don't get an answer to my question
 
Yes, that's how bounties work.
(although "entire reputation" is only true if you have very little reputation :P)
 
Ah, the Hadamard form
I should make an article on that, too
It's obscure enough
I got that big PDE book on the topic
I'd like to know where the fuck it comes from
 
Sid
2:40 PM
@ACuriousMind how will that be anymore helpful than asking another question specifically stating my confusion?
 
@Sid I already said that the primary goal is not to be maximally helpful to you, personally, and I don't claim that closing a question that asks the same question again saying why the existing answers are inadequate as a duplicate is helpful to the person who asks the question. But it is necessary in order to keep all the answers to the question in the same place.
 
Anonymous
@ACuriousMind That's better than casual searchers NOT finding an answer that suits them at all. More answers implies there's higher probability of finding an answer that suits them. From what I see (for duplicate questions):if they were left open they would receive more answers. But, once they are closed as duplicate hardly anyone goes back to the original question to add another answer. Preventing answers on duplicate questions just reduces the number of potential answers. That's worse.
 
@Slereah the one I checked out of the library that one time?
 
the very same
 
@Blue Are you suggesting that someone knows the correct answer and would post it to the new question, but upon seeing it's closed as a duplicate, turns away without adding their answer?
 
2:49 PM
@Blue Did you not see what ACM said? The SE model disregards what's best for the people.
 
Let's make our own Stack Exchange!
 
I should be proud because that's my own political philosophy, but...
 
@Slereah I'm certain BlackjackAndHookers.SE has been proposed :P
 
@Slereah Everyone who has tried to do that is dead. Coincidence?
 
Anonymous
@ACuriousMind Yes. I can say that from personal experience. And, I'm sure several others over here agree with that.
 
2:50 PM
That's because they weren't great!
 
@Blue But...why?
Is it out of a mistaken view of SE as a forum where old threads are dead and necroposting is an offense?
 
Yes
Also if it's old, there is less chances of getting points
 
@Slereah It still appears in the active tab and on the front page.
 
but the author of the post probably isn't around anymore
 
The only traffic you don't get is from those who only watch the "newest" tab, but idk who does that :P
Oh, okay, you probably won't get the potential +25 from OP, I'll grant you that
 
2:52 PM
@ACuriousMind SE is a forum. Why would you think otherwise
 
yeah, I don't even know what part of SE is actually regulars
Or if people on PSE mostly drop by once and ask a question to get an answer
 
@JohnRennie I'm going to disagree with you on that John. Even though your form groups all the type information together instead ("a is a pointer to int", nice) of grouping the fact that the variable is a pointer with the name instead of with the thing it points to ("an int is pointed to by a"< really?).
 
@0ßelö7 It's "a forum" in the sense of that it's a place where people come. But it's not a traditional internet forum in the message board tradition.
That's what people mean when they say "SE is not a forum", and you shouldn't pretend you don't know that :P
 
Because of the parallel with declaring global or local variables.
 
@dmckee these stylistic things inevitably come down to a matter of taste, and I realise that my taste differs from the mainstream in this respect.
 
2:55 PM
If you are in the habit of int* a you might write int* a, b; in a variable declaration and think that a and b have the same type.
 
@Slereah Most of the traffic is unregistered users. The proportion of accounts that are active only once to "regulars" should be extractable from SEDE
 
@JohnRennie Actually I think your taste sit he right one and the the language does something wrong to set up the possible confusion in multiple declarations.
 
But look at it this way. Suppose I use:
typedef int* pint;
then:
swap(pint a, pint b)
That's nice and clear.
 
The syntax ought to be <fully qualified type description> <list of variables that have that type>;
With no possibility of declaring both an int and a pointer-to-int in the same statement. Then we could use the better way to write it without risk of confusion.
It's a bug in c.
 
Given that pint and int* are the same type it seems logical to me to write swap(int* a, int* b)
Ah, our conversations overlapped - we agree! :-)
 
Anonymous
2:59 PM
@ACuriousMind Very few people would spend their time and energy answering a 1 year old question, when they could have answered a recent one instead. Answering a new question means they are helping a person who needs the answer soon. Answering an old question probably would not help the OP (of the old question) much. Also, you have less chances of getting your answer accepted by the OP when you answer old questions.
 
Anonymous
Agree or not, most answerers, knowingly or unknowingly, want to help the OP, in addition to adding to a database of knowledge. Think logically, which question would you want an answer to, right now? The one you posted 5 mins earlier, or the one you posted 5 years earlier? You probably don't need an answer to the question you asked 5 years earlier anymore!
 
Anonymous
SE really needs to take human emotions into account. The users are not robots.
 
@Blue I'm afraid that you're arguing architectual policy in a site moderation context. The place to take this up is on the mother meta.
But it's been hashed out repeatedly in the last ten years: answering duplicates has really bad consequences for the utility of the sites as repositories of knowledge.
 
Anonymous
@dmckee I intend to take it to meta. I was just replying to ACM
 
Anonymous
However, I don't think SE would implement it anyway.
 
3:02 PM
@Blue: I have some sympathy with your point of view, but I think there's a middle ground. If you make the question different enough to be not quite a duplicate I'll often answer it for exactly the reasons you describe.
 
And bad 'canonical' versions of the question are a problem with the current solution.
 
@Blue If I haven't posted an answer to the earlier question, I still want its answer. What you're saying is that most answerers don't use SE as it is intended to be used. That's fine, but not an argument to change the intention, it would be an argument for SE to come up with a better incentive to answer old neglected questions and turn them into the canonical sources they're supposed to be.
 
But if you allow unlimited exact duplicates we'll be swamped with a thousand twin paradox questions (instead of just several hundred as we have now :-)
 
People have been known to edit the original. And sometime a duplicate slips past the filters long enough to get good answer and surplant the earlier question as the new canonical one. Sometimes that's the best outcome.
 
3:04 PM
Which, by the way, is something I fully support - there should be more incentives for our users to polish old posts instead of answering the new incoming low-hanging fruit.
 
Anonymous
@ACuriousMind "it would be an argument for SE to come up with a better incentive to answer old neglected questions and turn them into the canonical sources they're supposed to be." I can agree with that
 
Anonymous
In fact I do agree with that. If that's implemented, well and good.
 
@Blue Of course, the main problem there is what that incentive could be.
 
Emilio is doing his bit in that regard by liberally dispensing bounties
 
^ Isn't he just.
 
3:05 PM
Maybe there could be a flag for bounty option, and if the question gets five flags a bounty is put on it.
 
@JohnRennie Thanks, I play whack-a-mole enough already with sockpuppets :P
 
(without anyone taking the rep hit)
 
@JohnRennie I haven't heard that suggestion before. You could take it up on meta.
 
The problem with most simple "solutions" is that they are horribly abusable by malicious users
 
I'm sure it would be a long time before any implementation as there is much to be thrashed out, but ... it might have merit.
 
3:07 PM
@dmckee I doubt the SE would consider it a priority
 
Anonymous
@ACuriousMind I'd just suggest giving out (by SE) bonus reps for answering well received questions older than 1 year and getting 1 or more upvotes on that answer. Something along that line...
 
@ACuriousMind you can never eliminate all abuse. The question is whether the benefits to genuine users outweigh the problems caused by the abusers.
 
@ACuriousMind The obvious requirement that the flag come from users with at least $N$ rep would suppress that. Especially if $N$ was, say, 5000.
 
@Blue Oh dear, can you imagine the flood of garbage answers just rephrasing the already present answers we would get? Your suggestion is excellent in a world where there are no people farming reputation.
 
Or maybe badge-holders in the tags on the question.
The idea would be to get expert users to agree that a question needs some better answers.
 
3:09 PM
@JohnRennie Bounties are (intended to be) very high visibility, though, so I'm rather wary of any automatic process that starts them
 
@Blue There are shiny badges for answering old questions, but consequential rewards open a really can of worms.
 
@dmckee Then we run into the opposite problem that not enough people will use that feature for it to really change anything...
 
Anonymous
@ACuriousMind Garbage answers wouldn't receive much upvotes I guess (I'd just suggest giving out (by SE) bonus reps for answering well received questions older than 1 year and getting 1 or more upvotes on that answer. b)
 
@dmckee The problem is that if the reward is small, it doesn't achieve much, and if it is large, it will attract all sorts of people trying to game it.
 
@ACuriousMind True. It certainly wouldn't work well on small site.
 
3:11 PM
I leave for 5 minutes and look what happens
What is going on
 
@Blue Number of upvotes is highly variable depending on the field the question is about. 3 upvotes for a high-traffic tag like newtonian-mechanics is not much, but in a low-traffic tag like conformal-field-theory it's already pretty good. So you might be generating an additional incentive for people to seek out high-traffic tags.
 
Anonymous
Duh, so many problems. Let's leave it as it is. :P
 
@Blue Well, that's essentially what happens with most ideas for features - it's pretty easy to determine the goal - "get people to answer old questions" - but it's pretty hard to come up with something doesn't also do something else.
That might also be the reason I can't really recall any major feature changes since I've become active here - the risk of breaking something "by accident" is rather high
 
Anonymous
Breaking things isn't really bad. You can't make better things till you break things. Anyhow, let's drop this for now.
 
"You can't make better things till you break things." - I think I've found a revolutionary.
3
 
Anonymous
3:20 PM
:P
 
3:46 PM
Yikes
We'd better get the drone ready
 
02:00 - 16:0016:00 - 00:00

« first day (2519 days earlier)      last day (1020 days later) »