16
$\begingroup$
  1. You may not use a calculator or computer.

  2. You may write "ln(X)" or "log(X)" to indicate the natural logarithm of X. Else, please let the reader know "log(X)" means log of X to the base 10, just to mention another common base.

  3. You are allowed to use $\ln(1 + x)\approx\ x-\dfrac{x^2}{2} + \dfrac{x^3}{3} \ $ for appropriate small values of $x$.

  4. To reduce some arithmetic, you are allowed to use these if they were to come up in calculations:

    $\ln(2) \ \approx \ 0.6931 $

    $\ln(3) \ \approx \ 1.0986 $

    $\ln(5) \ \approx \ 1.6094 $

  5. If it were to come up, you may use $\ \dfrac{\ln(5)}{\ln(4)} \ \approx \ 1.161$ in a calculation.


Please show the steps without using a calculator or computer to indicate which expression is larger:

$\large4^{5^9} \ \ \ \text{or} \ \ \ 5^{6^8}$

$\endgroup$
8
  • 3
    $\begingroup$ Challenging! The naive $4^7>5^6$ bound is not tight enough! $\endgroup$ Commented Dec 17, 2024 at 21:36
  • 6
    $\begingroup$ I computed ln(4)*(5^9) and ln(5)*(6^8), they are within ~0.16% of each other! $\endgroup$
    – Florian F
    Commented Dec 18, 2024 at 7:57
  • 3
    $\begingroup$ @FlorianF The logs are close, but the numbers themselves aren't particularly close; one is more than 10^1897 times bigger than the other! $\endgroup$ Commented Dec 18, 2024 at 18:23
  • 1
    $\begingroup$ A method that does not work: Let $f(x)= x^((x+1)^(2x+1))$ and let $g(x)= (x+1)^((x+2)^(2x)).$ Then one might hope that one of these is always bigger than the other. But if one compares $\ln \ln f(x) = \ln \ln x + (2x+1) \ln (x+1)$ to $\ln \ln (x+1) + 2x \ln (x+2)$ one will see this doesn't work. And if one does cheat and use a calculator, one will find that the two cross at $x=3.98$ which helps explain why this is a genuinely difficult quantity to estimate, since the relevant $x$ value here is $x=4$. $\endgroup$
    – JoshuaZ
    Commented Dec 18, 2024 at 23:30
  • 3
    $\begingroup$ @RosieF re "can't practicably be done without a computer". My original answer was a scan of a hand-written solution, which did not use a computer or calculator, and Benjamin rewrote it as MathJax text. This is a puzzle site, and my "aha" moment was seeing how it could be done quite easily by hand, using only one of the several pieces of information given. $\endgroup$ Commented Dec 19, 2024 at 11:28

8 Answers 8

18
$\begingroup$

Working by hand, I can show that

$$\begin{matrix}5^1 & 5 & 6^1 & 6 \\5^2 & 25 & 6^2 & 36 \\5^3 & 125 & 6^3 & 216 \\5^4 & 625 & 6^4 & 1296 \\5^5 & 3125 & 6^5 & 7776 \\5^6 & 15625 & 6^6 & 46656 \\5^7 & 78125 & 6^7 & 279936 \\5^8 & 390625 & 6^8 & 1679616 \\5^9 & 1953125 & & \end{matrix}$$

So $5^9 = 1953125$ and $6^8 = 1679616$

Compare

$$4 ^ {5 ^ 9} \text{ vs } 5 ^ {6 ^ 8} $$ $$4^{1953125} \text{ vs } 5^{1679616}$$ $$ 1953125\times\ln{4} \text{ vs } 1679616\times\ln 5$$ $$\frac{1953125}{1679616} \text{ vs } \frac{\ln{5}}{\ln{4}}$$

Divide longhand to show that this is true

longhand latex screenshot made with longdivision package

The longhand division shows $$\frac{1953125}{1679616} > 1.162$$ and, when knowing $\frac{\ln{5}}{\ln{4}}$ is 1.161 to 3 decimal places, means $$\frac{1953125}{1679616} > \frac{\ln{5}}{\ln{4}}$$ and thus $$4 ^ {5 ^ 9} > 5 ^ {6 ^ 8} $$

$\endgroup$
0
12
$\begingroup$

The claim to check is

$$4^{5^9} \stackrel{?}{>} 5^{6^8}$$

We will check it without logarithms.

Take the $5^8$th root of both sides to get the equivalent:
$$4^{5} \stackrel{?}{>} 5^{1.2^8}$$
We can calculate ${1.2^8}$ by hand by squaring 3 times: $1.2 \to 1.44 \to 2.0736 \to 4.29981696$. This is just below $4.3$

So it is enough to check

$$4^{5} \stackrel{?}{>} 5^{4.3} (> 5^{1.2^8})$$ which is equivalent to $$4^{50} \stackrel{?}{>} 5^{43}$$ which is equivalent to $$2^{143} \stackrel{?}{>} 10^{43}$$ which is equivalent to $$(2^{10})^{14} \cdot 2^3 \stackrel{?}{>} (10^3)^{14} \cdot 10 $$

Since $2^{10}=1024$, it is equivalent to

$$1.024^{14}\cdot 8 \stackrel{?}{>} 10$$ which is true because $(1+x)^n > 1+nx$ gives $$1.024^{14} > 1+14\times 0.024 = 1.336 > 1.25$$

So,

$$4^{5^9} > 5^{6^8}.$$

$\endgroup$
1
  • 1
    $\begingroup$ Thank you for editing @BenjaminWang It is much better now (and I learned some new layout notation) $\endgroup$
    – Retudin
    Commented Dec 19, 2024 at 15:44
7
$\begingroup$

Starting with the claim

$$ 4^{5^9} > 5^{6^8} $$

We have

$$ \iff 5^9\ln4 > 6^8\ln5 $$ $$ \iff \frac{5^9}{6^8} > \frac{\ln5}{\ln4} $$ $$ \iff 5\times\left(\frac56\right)^8 > \frac{\ln5}{\ln4} $$ $$ \iff \ln5+8\ln\frac56 > \ln\frac{\ln5}{\ln4} $$

Using $\ln x = - \ln(1/x)$, we can define a quantity $A$:

$$ \iff A:= \frac{\ln5 - \ln\frac{\ln5}{\ln4}}{\ln\frac65} > 8 $$

Using the hints we have the bounds

$$ 0.69305 < \ln(2) < 0.69315$$ $$ 1.09855 < \ln(3) < 1.09865$$ $$ 1.60935 < \ln(5) < 1.60945$$

Since we want to prove that $A>8$, we have to try to make it smaller using our bounds while still having the result be larger than $8$.

For the first term in the numerator, if we make it smaller, then $A$ will get smaller, so we need $$\ln(5) > 1.60935.$$ For the second term in the numerator, if we make it bigger, then $A$ will get smaller, so we can first do longhand division to get $$ \frac{\ln5}{\ln4} = \frac{\ln5}{2\times\ln2}< \frac{1.60945}{2\times 0.69305} < 1.1612 $$ (the given hint only upper bounds it by 1.1615 which will not be good enough), then put this in the series to compute $$\ln\frac{\ln5}{\ln4} < \ln 1.1612 < 0.1612 - \frac{0.1612^2}{2} + \frac{0.1612^3}{3}$$ (it is an upper bound since we have just added a positive term, the cubed term, to an alternating series). We can get an upper bound on this by only ever keeping 4 decimal places. Let $x=0.1612$ then compute $0.0258 < x^2 < 0.0260$. Use the upper bound for $x^2$ to compute $x^3 < 0.0260x < 0.0042$, so this is $$<0.1612 - 0.0129 + 0.0014 = 0.1497$$ For the denominator, if we make it bigger, then $A$ will get smaller, so we need $$ \ln(6/5) = \ln2 + \ln3 - \ln5 < 0.69315 + 1.09865 - 1.60935 = 0.18245 $$

we get that the claim

$$ \iff \frac{\ln5 - \ln\frac{\ln5}{\ln4}}{\ln(6/5)} > \frac{1.60935 - 0.1497}{0.18245} = 8 + \frac{1}{3649}> 8$$

which is true

$\endgroup$
5
  • $\begingroup$ At the end of the last box you have the left-hand side is less than the right-hand side, but you started with the claim that the left-hand side is greater than the right-hand side. $\endgroup$ Commented Dec 18, 2024 at 3:22
  • 1
    $\begingroup$ The inequality reverses because both sides are divided by a negative value: the logarithm of a number less than 1. $\endgroup$ Commented Dec 18, 2024 at 5:14
  • $\begingroup$ I hope you don't mind my edits. Explanation: to get the correct bound you need to add or remove (depending on where it appears in the inequality) $0.5$ times the limit of precision. When trying this, I found that the original strategy leads to bounds that are not tight enough. I fixed this by using $\ln5$ (given) instead of $\ln6$ (has to be calculated). As you can see, it is very close: we only managed to prove by one part in $8 \times 3649 = 29192$. This is not surprising, as $\ln\ln 4^{5^9} \approx 14.81157547 < 14.80996 \approx \ln\ln 5^{6^8}$, which differ in only one part in over $9000$. $\endgroup$ Commented Dec 18, 2024 at 13:27
  • $\begingroup$ @BenjaminWang,many thanks for your edits, I'm happy to see this presented in a better way! $\endgroup$
    – ThomasL
    Commented Dec 18, 2024 at 22:47
  • 1
    $\begingroup$ @Benjamin Wang - You have a typo in your second to last line. It should be ... 14.81157547 > 14.80996 ... $\endgroup$ Commented Dec 19, 2024 at 14:44
3
$\begingroup$

Let us first try to reduce the exponents to manageable sizes. The idea is to find positive integers $x$ and $y$ such that $4^{5^9 - x} > 5^{6^8 + y}$ while $\text{gcd}(5^9 - x, 6^8 + y)$ is large. We can find that the largest gcd is $54253$, achieved when $x = 17$ and $y = 2227$. Now it suffices to show that $$ 4^{5^9 - 17} > 5^{6^8 + 2227} \Leftrightarrow 4^{54253 \cdot 36} > 5^{54253 \cdot 31} \Leftrightarrow 4^{36} > 5^{31}$$ The numbers are a quite big but it's possible to compute by hand: \begin{align*} 4^{36} &= 4\,722\,366\,482\,869\,645\,213\,696 \\ 5^{31} &= 4\,656\,612\,873\,077\,392\,578\,125 \end{align*} From here we can see that $4^{36} > 5^{31}$ and therefore $4^{5^9} > 5^{6^8}$.

$\endgroup$
2
  • $\begingroup$ Very nice approach! And welcome to PSE :) $\endgroup$ Commented Dec 20, 2024 at 15:04
  • 2
    $\begingroup$ How did you find $54253$? Computing $5^9$ and $6^8$ by hand is easy enough, but how would you go about finding a large divisor close to both? $\endgroup$
    – JiK
    Commented Dec 20, 2024 at 15:11
3
$\begingroup$

For starters,

We wish to show that $4^5 > 5^{1.2^8}$, as raising both sides to the power $5^8$ will resolve the initial inequality, so we will need to prove that $\frac{5\ln(4)}{\ln(5)} > 4.306 > 1.2^8$.
The first inequality follows from the given precisions for $\ln(2)$ and $\ln(5)$, even if we were given the worst combination of rounding errors. Simply using $\frac{5}{1.161}$ would give us $4.3066$ to work with instead.

Time to dust off an old mathematical trick!

Using the series for $\ln(1+x)$ whose first few terms have been given to us, subtract from it a copy with $x$ negated to yield $\ln\left(\frac{1+x}{1-x}\right)= 2\left(x+\frac{x^3}{3}+\frac{x^5}{5}+\dots\right)$. Substituting $\frac{1}{2n+1}$ yields a particularly useful form:
$1.) \ln\left(1+\frac{1}{n}\right) = \frac{2}{2n+1}+\frac{2}{3(2n+1)^3}+\frac{2}{5(2n+1)^5}+\dots$
We can get a strong upper bound by replacing coefficients higher than 3 with 3, and using the standard rule of geometric series sums to get
$2.) \ln\left(1+\frac{1}{n}\right) < \frac{2}{2n+1}+\frac{1}{6n(n+1)(2n+1)}$

We can apply this rule two different ways to show the final inequality:

Using 2.) with $n=5$ yields $8\ln(1.2) < \frac{16}{11}+\frac{2}{495} < 1.4586$.
Using 1.) with a single term, we have that $\ln(4.306) = \ln(4)+\ln(1.0765) > 1.3862+\ln\left(1+\frac{1}{13.1}\right) > 1.3862+\frac{1}{13.6}> 1.4597$.
And we are done!

$\endgroup$
4
  • $\begingroup$ I don't know how you conclude/demonstrate that inequality in the top left of your top box, and how it directly relates to the main inequality. $\endgroup$ Commented Dec 18, 2024 at 3:35
  • $\begingroup$ It's the statement I wish to prove, which is the 5^8'th root of the main inequality. $\endgroup$ Commented Dec 18, 2024 at 5:11
  • 1
    $\begingroup$ You might want to make that more obvious; as is, it looks like you're asserting it. $\endgroup$
    – Idran
    Commented Dec 18, 2024 at 15:42
  • $\begingroup$ I think you mean substituting $\frac{1}{2n+1}$ for $x$, not $2n+1$. $\endgroup$
    – YurkoFlisk
    Commented Dec 19, 2024 at 16:55
2
$\begingroup$

This answer begins similarly to ThomasL's, but diverges partway through.

We want to know if $4^{5^9} > 5^{6^8}$. That's the same as asking whether $$ \ln{4^{5^9}} \stackrel{?}{>} \ln{5^{6^8}}$$ $$ 5^9 \ln{4} \stackrel{?}{>} 6^8 \ln{5} $$ or $$ \frac{5^9}{6^8} \frac{\ln{4}}{\ln{5}} \stackrel{?}{>} 1 $$

Since we are given

$$ \frac{\ln{5}}{\ln{4}} \approx 1.161 $$

we need only determine whether

$$ \frac{5^9}{6^8} \stackrel{?}{>} 1.161 $$

Now I can compute by hand

$$ \begin{aligned} \ln{\frac{5^9}{6^8}} & = 9 \ln 5 - 8\left(\ln 3 + \ln 2\right) \\ & = 9 \times 1.6094 - 8 \times (1.0986 + 0.6931) \\ & = 14.4846 - 14.3336 \\ & = 0.1510 \end{aligned}$$ If this were negative, I'd be able to stop, because then we'd know $5^9 < 6^8$, and its product with $\ln 4 / \ln 5$ would be smaller than unity.

However, I can make use of the fact that

$$ e^x = 1 + x + \frac{x^2}{2} + ... $$

to work out that

$$ \begin{aligned} \frac{5^9}{6^8} \approx e^{0.1510} & \approx 1 + 0.1510 + 0.5 \times (0.1510)^2 + \ldots \\ & = 1.1510 + 0.5 \times 0.022801 + \ldots \\ & = 1.1624 + \ldots \end{aligned}$$ This is already larger than the 1.161 I'm comparing it to, so I don't need to compute any more terms in the expansion; they're all positive so they will only increase the sum.

Thus

$$ \frac{5^9}{6^8} \frac{\ln 4}{\ln 5} > 1 $$

and so

$$ 4^{5^9} > 5^{6^8} $$

$\endgroup$
8
  • $\begingroup$ Welcome to Puzzling.SE. This is a solid attempt, but unfortunately this suffers the same pitfall as ThomasL's original answer, which is that approximations are not inequalities. Please check my comment under his answer. $\endgroup$ Commented Dec 19, 2024 at 9:19
  • $\begingroup$ @chenel - The main issue I had not foreseen is that I did not give my approximations in the statement of my problem more correct decimal digits. For instance, the multiplying by 9 and the multiplying by 8, respectively, compound the round off errors. It makes the calculations and comparisons after that untrustwothy. $\endgroup$ Commented Dec 19, 2024 at 15:16
  • $\begingroup$ Hmm. I'm not sure the logic actually holds up in this particular case. Let me assume that all of ln 2, ln 3, and ln 5 should instead be rounded oppositely in the way that's maximally infavorable for my method. Then I'm computing instead 9×1.6093?8×(1.0987+0.6930) = 0.1501 and my exponential is 1.1501 + 0.1501^2/2 = 1.1619. That's the lower bound on my estimate. So long as the given ln 5 / ln 4 ratio is not rounded incorrectly, this method still holds up, no? $\endgroup$
    – chenel
    Commented Dec 19, 2024 at 15:28
  • 1
    $\begingroup$ Since you want to prove that $9\ln 5 -8(\ln3+\ln2)$ is greater than something, to make the approximations maximally unfavorable to you you need to shrink the positive terms and expand the negative terms (by 0.5 times the limit of precision). (I think you accidentally shrunk the $\ln2$ term) This means it is $$>9(1.60935)-8(1.09865+0.69315) = 0.14975=:x.$$ Now using $e^x > 1+x+x^2/2=1.1609$ is insufficient. Fortunately, adding one more term $x^3/6$ gets you over $1.1615$, which is more than $\ln 5 / \ln 4$ which is $1.161$ to 3 decimal places. $\endgroup$ Commented Dec 20, 2024 at 0:00
  • 1
    $\begingroup$ @OliveStemforn your hints are fine (more decimal places means that doing $x^3$ is even more infeasible without a calculator). The problem itself is what's difficult (it is a very tight inequality). Perhaps it was better for the Mathematics StackExchange. $\endgroup$ Commented Dec 20, 2024 at 0:26
1
$\begingroup$

Originally I wanted to avoid relying on any of the logarithm approximations, unfortunately the results are just close enough that we must rely on one of them. This solution seeks to use methods that make by-hand calculations easier to reason about, due to seeing some comments saying it's much too hard to calculate by hand.

$$4^{5^9} \stackrel{?}{>} 5^{6^8}$$ $$5 \cdot 5^8 \cdot 2 \ln(2) \stackrel{?}{>} 6^8 \cdot \ln(5)$$ $$10 \frac{\ln(2)}{\ln(5)} \stackrel{?}{>} \left ( \frac{6}{5} \right )^8$$ $$\frac{\ln(2)}{\ln(5)} \stackrel{?}{>} 10^{-1} \cdot 1.2^8$$

Now, let's simplify that right hand side, it's numeric so we can just calculate it but I'll try and keep it as "easy" math. I assume the reader knows one of the following, though they're not too hard to quickly do by hand: $2^8 = 256$ or $2^{10} = 1024$ (which can be quickly derived from $2^8$)

$$\Rightarrow 1.2^8 = 10^{-8} \cdot 12^8$$ $$\Rightarrow 12^8 = 4^8 \cdot 3^8 = 2^{16} \cdot 3^8$$ $$\Rightarrow 3^8 = 9^4 = 81^2 = \left(80 + 1\right)^2 = 80^2 + 2\cdot 80 + 1 = 6400 + 160 + 1 = 6561$$ $$\Rightarrow 2^{16} = 2^{6} \cdot 2^{10} = 4 \cdot 4 \cdot 4 \cdot 1024 = 4 \cdot 4 \cdot 4096 = 4 \cdot 4(4100 - 4) = 4 \cdot 16384$$ $$\Rightarrow 1.2^8 = 10^{-8} \cdot 4 \cdot 10^{4} \cdot 1.6384 \cdot 10^4 \cdot 6.561 $$ $$\Rightarrow 4 \cdot 6.561 = 26.244 = 10 \cdot 2.6244$$ Meaning we can now simplify to $$\frac{\ln(2)}{\ln(5)} \stackrel{?}{>} 10^{-1} \cdot 1.2^8 = 10^{-1} \cdot 1.6384 \cdot 2.6244$$ Unfortunately, the numbers are too close to each other for us to use simple math or approximations anymore... Let's first get those logs, via the given for $\log_4(5) \approx 1.161$, noting that across multiplication we need to round up (it could be anywhere between 1.1610 and 1.1619) $$ \frac{\ln(2)}{\ln(5)} = \frac{1}{2 \cdot \frac{\ln(5)}{\ln(4)}} = \frac{1}{2 \cdot 1.161} = \frac{1}{2.323}$$ Inverting the expression we have $$ 2.323 \stackrel{?}{<} \frac{10}{1.6384 \cdot 2.6244}$$ As nice as it would be to just approximate to 1.7 times 2.7 or 1.64 times 2.63, we need to multiply up to 1.639 times 2.625 at least for precision. Note, it's rounded up to guarantee the result is smaller than the true value (since it's a fraction). $$\Rightarrow 1.639 \cdot 2.625 = 10^{-6} \cdot (1640 - 1) \cdot 2625$$ $$\Rightarrow 1640 \cdot 2625 = 10 \cdot 41 \cdot 4 \cdot 2625 = 10 \cdot 41 \cdot 10500 $$ $$ \Rightarrow 10^3 \cdot (40 + 1) \cdot 105 = 10^3 \cdot (4200 + 105) = 10^3 \cdot 4305 $$ $$ \Rightarrow 10^{-6} \cdot (10^3 \cdot 4305 - 2625) = 10^{-6} \cdot (4305000 - 2625) = 10^{-6} \cdot 4302375$$ Okay, just one more division to go. Trial division with that large number is pretty painful, but thankfully taking it as 4.3024 gives us enough precision, again rounding up. $$ 2.323 \stackrel{?}{<} \frac{100000}{43024} $$ To illustrate long division I will just list multiplications and remainders $$ \frac{100000}{43024} = \frac{86048 + 13952}{43024} = 2 + 10^{-1} \frac{129072 + 10448}{43024} = 2 + 0.3 + 10^{-2} \frac{86048 + 18432}{4303}$$ Our last remainder is above $4 \cdot 43024 = 172096$ so we have $$ \frac{100000}{43024} \geq 2 + 0.3 + 0.02 + 0.004 = 2.324 $$ Which means we have $ 2.323 \stackrel{?}{<} 2.324 $, which is true. Therefore, our original statement is true and $4^{5^9} > 5^{6^8}$.

$\endgroup$
-1
$\begingroup$

Let $f=x^u$. If we take the derivative with respect to $x$, we get $df =ux^{u-1}dx$, which we can write as $$df =\frac u x f dx. \quad \cdots (1)$$ If we take the derivative with respect to $u$, we get $df=(\ln u) f du \cdots (2)$.

Now suppose we have $u=y^z$. We then get $du = \frac z y u dy$ and $du = (\ln z ) u dz$. If we plug those into $(2)$, we get (respectively) $$df = (\ln u) \frac z y uf dy \cdots (3)$$ and $$df = (\ln u)(\ln z ) ufdz \quad \cdots(4)$$

In $(1)$, $(3)$ and $(4)$ we can first disregard the common factor $uf$, and get $$\frac 1 x dx \quad \text{ vs } \quad (\ln u) \frac z y dy \quad \text{ vs } \quad (\ln u)(\ln z)dz.$$ The first one, $\frac 1 x dx$, is clearly the smallest of these, so we can then move on to compare the last two. These two have the common factor $(\ln u)$, and removing that leaves us with $$\frac z y dy\quad \text{ vs } \quad \ln z dz$$

If we plug in $x=4, y=5, z=9$, we get $\frac z y =1.8$ and $\ln z=2.19$. So $\ln z > \frac z y$.

This hold all along the way going to $x=5, y=6, z=8$. $z$ is decreasing and $y$ is increasing, so $\frac z y$ is decreasing. $\ln z$ monotonically decreases as $z$ decreases, so its minimum is at $z=8$. $\ln 8 =3\ln3 = 2.079$. The minimum value of $\ln z$ is greater than the maximum value of $\frac z y$, thus $\ln z > \frac z y$.

Going from $x=4, y=5, z=9$ to $x=5, y=6, z=8$, we have that $dx$ and $dy$ are positive, and $dz$ is negative, so $df$ is dominated by a decrease, and hence $f$ is decreasing.

$\endgroup$
2
  • 1
    $\begingroup$ Let $f(x,y,z)=x^{y^z}$. You found its gradient $$\nabla f(x,y,z)=\left(\frac u x f, uf\ln u\frac z y, uf\ln u\ln z\right)$$ where $u =y^z$ and $f = x^u$. And then you argued that $\nabla f(x,y,z)$, locally at $a:=(4,5,9)$, in the rough direction of $v:= \begin{pmatrix}1\\1\\-1\\\end{pmatrix}$, being $q:=(\nabla f(a)) v$, is $<0$. Even if this is true, it is not enough, because gradient is a local property. (What if $q>0$ for some other $a$?) I suppose you can try to show that it is decreasing on the entire line (or some path) between $(4,5,9)$ and $(5,6,8)$, but I imagine that might be hard. $\endgroup$ Commented Dec 21, 2024 at 10:09
  • $\begingroup$ @BenjaminWang It seems rather trivial to me. $\endgroup$ Commented Dec 21, 2024 at 19:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.