![\""Meow"\"](\"/public/images/Pasted%20image%2020260622201004.png\")
I had a loop that walked a vector backwards and it never terminated. The bug is\nthat size_t is unsigned, so it can never be negative — subtracting past\nzero wraps to a huge positive number instead.
// Broken: when i == 0, --i wraps to SIZE_MAX, so i >= 0 is always true.\nfor (size_t i = v.size() - 1; i >= 0; --i) {\n use(v[i]);\n}\nThe same trap hides in any v.size() - 1 when the vector is empty: 0u - 1\nis SIZE_MAX, not -1, so a \"last index\" computed that way points way off the\nend.
One fix is to offset the index by one and count the other variable down, so\nthe loop variable stays in valid unsigned territory:
\n// Fixed: i is the count remaining; the real index is i - 1.\nfor (size_t i = v.size(); i > 0; --i) {\n use(v[i - 1]);\n}\nThe general rule: if a value can go negative, don't store it in an unsigned\ntype. Reach for a signed index (or C++20's std::ssize) when you need to count\nbelow zero.
This site is a running log of things I learn and the projects I build. The most\nimportant property it has is that publishing is frictionless: writing a new post\nis exactly as much work as creating a Markdown file and pushing to git. No CMS,\nno dashboard, no database.
\nIt's built with Astro as a fully static site. Every TIL\nand every post is a Markdown file with a little frontmatter block at the top —\na title, a date, some tags. Astro reads those files at build time, validates the\nfrontmatter against a schema, and turns each one into a page. The output is\nplain HTML and CSS with essentially no client-side JavaScript, so pages load\ninstantly and there's nothing to break.
\nThere are two kinds of writing here:
\nBecause the content is the product. A personal site that's mostly text doesn't\nneed a rendering framework shipped to the browser; it needs good typography, a\nreadable column width, and fast loads. Keeping the build static means the whole\nsite is a folder of HTML files that any host can serve, and there's no runtime\nto patch or attack.
\nEverything is available as a feed. There's a\ncombined feed of TILs and posts, and if you only want one or the\nother, each has its own. Feeds carry the full text, so you can read without ever\nvisiting the site — that's the point of them.
\nIf you want to see how the sausage is made, the source is on\nGitHub.
\n","date_published":"2026-07-03T00:00:00.000Z","tags":["meta","astro"]},{"id":"https://shreyashrai.com/links/the-original-link-blog/","url":"https://shreyashrai.com/links/the-original-link-blog/","title":"Simon Willison's link blog — the format worth stealing","content_html":"Kicking off a link blog here, the way Simon Willison\ndoes it: short posts that point at something worth your time with a line on why.\nLess pressure than an essay, and over a year it becomes the most useful thing on\nthe site. This is the first one.
\nLink: https://simonwillison.net/
","date_published":"2026-07-02T00:00:00.000Z","tags":["meta","blogging"]},{"id":"https://shreyashrai.com/quotes/willison-on-writing/","url":"https://shreyashrai.com/quotes/willison-on-writing/","title":"Quote — Simon Willison","content_html":"You should blog about things you've learned, even if you don't think anyone will\nread it. The act of writing it up is what cements your own understanding.
\n","date_published":"2026-07-01T00:00:00.000Z","tags":["blogging","meta"]},{"id":"https://shreyashrai.com/links/astro-content-layer/","url":"https://shreyashrai.com/links/astro-content-layer/","title":"Astro's Content Layer API","content_html":"The glob() loader plus Zod schemas are what make every TIL, post, link, and\nquote on this site type-checked at build — a malformed frontmatter block fails\nastro build instead of shipping a broken page. Worth a read if you run any\nkind of content site.
Link: https://docs.astro.build/en/guides/content-collections/
","date_published":"2026-06-28T00:00:00.000Z","tags":["astro","meta"]},{"id":"https://shreyashrai.com/til/linked-lists/","url":"https://shreyashrai.com/til/linked-lists/","title":"Linked Lists","content_html":"Array has a fixed size, so partially filled array takes as much space as it was full.
\nSo there's a lot of unused space in memory.
\nSometimes, memory may not be available all the time since contiguous memory so it requires copying of data.
\nLinked List requires extra memory for pointer variables. They are 4 bytes each (in 32bit system). But 8 bytes in 64bit system.
\nBut it does not leave unused memory.
\nIt's usually better than array when the datatype being stored has higher size. But depends on the case on which performs better.
\nNodes being stored at random places in memory.
\nMaking a node
\ntypedef struct Node{\n\tint data;\n\tstruct Node* next;\n} Node;\nstruct Node{\n\tint data;\n\tNode* next;\n};\nInitializing Linked List
\nNode* A; //initialising empty LL\nA = NULL; //Setting it as Null initially\nNode* temp = (Node*) malloc(sizeof(struct Node)) //typecasting required, creating a new temporary node\n(*temp).data = 2; //Filling data in node\n(*temp).next = NULL; //Attaching it to null address\nA = temp; //Setting A as temp node\nNode* A;\nA = NULL;\nNode* temp = new Node();\ntemp->data = 2; // -> is basically short form of the deferencing (*) and accessing structure with .\ntemp->next = NULL;\nA = temp;\nNode* temp1 = A;\nwhile (temp->next != NULL){\n\ttemp1 - temp1 -> next;\n}\n\ntemp = new Node();\ntemp->data = 4;\ntemp->next = NULL;\ntemp1->next = temp;\n#include <stdlib.h>\n#include <stdio.h>\n\ntypedef struct Node {\n\tint data;\n\tstruct Node* next;\n} Node;\n\nvoid Insert(Node** pointerToHead, int x);\nvoid Print(Node* head);\n\nint main(){\n\tstruct Node* head = NULL;\n\tint n,x;\n\tprintf(\"Enter number of nodes: \")\n\tscanf(\"%d\", &n);\n\tfor (int i=0; i < n; i++){\n\t\tprintf(\"Enter a number: \");\n\t\tscanf(\"%d\", &x);\n\t\tInsert(&head, x);\n\t\tPrint(head);\n\t}\n\tprintf(\"\\n\");\n}\n\nvoid Insert(Node** pointerToHead, int x){\n\tNode* temp = (Node*) malloc(sizeof(struct Node));\n\ttemp->data = x;\n\ttemp->next = *pointerToHead;\n\t*pointerToHead = temp;\n}\n\nvoid Print(Node* head){\n\twhile (head != NULL){\n\t\tprintf(\"%d \",head->data);\n\t\thead = head->next;\n\t}\n\tprintf(\"\\n\");\n}\nNotes:
\nPremature optimization is the root of all evil.
\n","date_published":"2026-06-20T00:00:00.000Z","tags":["programming"]}]}