I want to share the resources that helped me set up this ghost blog. First comes the domain name. I did some research and I got the .com domain name I wanted at a great price from porkbun, 29.75₹ per month (5$ for a year)!

Now you should expect to fork out 600₹+ every year to renew these domains. The first two years are usually the cheapest.

Now self hosting (using your personal computer as a server) is too complicated and I would not recommend it. You can apply for the github student pack and use the credits for digital ocean. Then it's a matter of a single click to deploy your droplet. But the 50$\$$ github gives you and maybe 100$\$$ that you can get by using some referal code online won't last you beyond 3 months.

I recommend google. For a year you have 21,343₹ to spend. After that you can start spending real money. You can always backup ghost then and go to a cheaper hosting.

img
$300 free credits as soon as you give them your card details

Navigate to Compute Engine -> VM instances -> Create an instance. Then search the marketplace for Bitnami Ghost which is the cheapest.

img
971.82₹ is a bit much but let's see after a year. 11,661 < 12k

You should then follow the instructions given, make your server IP static. Point your domain to that IP (you can create a A record in the porkbun site). SSH into your server and enable https using Let's Encrypt. You will want to remove the irritating banner at the bottom right corner.

Now the main reason I chose ghost over just blogging on Medium is because I want support for latex. It is so stupid how they don't let us enable it. I guess they don't want the overhead at that scale. You can insert the code shared by feynman and voila! That code is now deprecated, we need to make a small change to that as described here.

There you have $\vec x \in \mathbb{R}^n$ and $\mathscr{y}$. Or maybe when you want to say $\exists \ p \in \mathbb{Q}$ such that,

$$a < p < b$$

Where $a, b \in \mathbb{R}$

$$\sum_{n = 0}^{\infty} 10\times (0.5)^{n} = \frac{10}{1-\frac{1}{2}}$$

Well it is not just math that I can write now. I can write pretty code!

def hilbert(n):
    h = np.zeros((n,n))
    for i in range(n):
        for j in range(n):
            # Since i and j are indexed from 0 while math indexes from 1
            h[i][j] = 1/((i+1)+(j+1)-1)
            
    return h
A famously ill-conditioned problem. Hilbert Matrix.

After your ghost is ready, you will want to create a nice transparent logo for your blog. I used a white background and edited it to have a text over it, then once it's cropped to the required aspect ratio use an online tool to make the white part transparent.

Well, ask away in the comments or contact me personally if you face any issues.