${esc(e.title)}

[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_217":218,"_219":220},"routes/_app.x.$handle.$name",{"_103":102,"_32":45,"_111":90,"_27":112,"_13":14,"_113":-5,"_11":114,"_115":116,"_117":118,"_119":120,"_121":122,"_20":-5,"_123":124,"_125":126,"_127":25,"_128":129,"_130":59,"_131":-5,"_132":-5,"_133":134},"routes/_app.x.$handle.$name.branch.$branch.version.$version.diff",{"_11":12,"_13":14,"_15":16,"_17":-5,"_18":-5},"project",{"_43":44,"_32":45,"_38":46,"_47":-5,"_48":25,"_49":50,"_51":50,"_52":53,"_54":-5,"_18":55,"_56":57,"_58":59,"_60":-5,"_61":-5,"_62":63,"_64":65,"_66":67,"_68":69,"_70":71,"_72":-5,"_73":74,"_75":74,"_76":77,"_78":77,"_79":77,"_80":74,"_81":74,"_82":83,"_84":85,"_86":87,"_88":25},"version",2,"changes",[19],"aiChatInfo","userId",{"_20":21,"_22":23,"_24":25,"_26":25,"_27":28,"_29":28},"file",{"_30":31,"_32":28,"_33":34,"_35":37,"_38":39,"_40":14,"_41":42,"_24":25},"previous",{"_30":31,"_32":28,"_33":34,"_35":36},"deleted",false,"deletedHere","path","scraper.ts","latestPath","fileId","01a055cb-fa00-72a2-9988-e9c9b521e580","name","fileType","script","content","import { blob } from \"https://esm.town/v/std/blob/main.ts\";\n\nexport const BASE = \"https://paulgraham.com\";\n\n// Blob key holding the cached feed data. Using blob storage (not SQLite) keeps\n// the feed independent of the libsql client, which has been flaky on Deno.\nconst KEY = \"pg-essays-cache\";\n\ninterface CacheEntry {\n title: string;\n date: string | null;\n position: number;\n}\n\ninterface Cache {\n lastCheck: string | null;\n essays: Record;\n}\n\n// Files on paulgraham.com that are in articles.html but aren't essays.\nconst EXCLUDE = new Set([\n \"/index.html\",\n \"/bio.html\",\n \"/rss.html\",\n \"/rss.xml\",\n \"/feed.xml\",\n]);\n\nasync function readCache(): Promise {\n const c = await blob.getJSON(KEY);\n return c ?? { lastCheck: null, essays: {} };\n}\n\nasync function writeCache(c: Cache): Promise {\n await blob.setJSON(KEY, c);\n}\n\nasync function fetchIndexHtml(): Promise {\n const res = await fetch(`${BASE}/articles.html`);\n if (!res.ok) throw new Error(`articles.html: HTTP ${res.status}`);\n return await res.text();\n}\n\n/** Parse the ordered (newest-first) list of essay links from the index. */\nexport function parseIndex(html: string): { title: string; url: string }[] {\n const out: { title: string; url: string }[] = [];\n const re = /]*>([\\s\\S]*?)<\\/a>/gi;\n let m: RegExpExecArray | null;\n while ((m = re.exec(html)) !== null) {\n let href = m[1];\n const text = m[2].replace(/<[^>]+>/g, \"\").replace(/\\s+/g, \" \").trim();\n if (href.startsWith(BASE)) href = new URL(href).pathname;\n // Normalize bare relative hrefs (e.g. \"greatwork.html\") to \"/greatwork.html\".\n if (!href.startsWith(\"/\")) href = \"/\" + href.split(/[?#]/)[0];\n const clean = href.split(/[?#]/)[0];\n if (!clean.endsWith(\".html\")) continue;\n if (EXCLUDE.has(clean)) continue;\n if (!text) continue;\n out.push({ title: text, url: clean });\n }\n const seen = new Set();\n return out.filter((x) => (seen.has(x.url) ? false : (seen.add(x.url), true)));\n}\n\n/**\n * Best-effort publication date from Hacker News: the earliest paulgraham.com\n * submission of the essay. Returns an ISO string or null.\n */\nasync function hnDate(title: string): Promise {\n try {\n const q = encodeURIComponent(title);\n const res = await fetch(\n `https://hn.algolia.com/api/v1/search?query=${q}&tags=story&hitsPerPage=50`,\n );\n if (!res.ok) return null;\n const data = await res.json();\n const hits: any[] = data.hits ?? [];\n let min: number | null = null;\n for (const h of hits) {\n let host = \"\";\n try {\n host = h.url ? new URL(h.url).hostname : \"\";\n } catch {\n host = \"\";\n }\n if (host === \"paulgraham.com\" || host.endsWith(\".paulgraham.com\")) {\n if (typeof h.created_at_i === \"number\" &&\n (min === null || h.created_at_i < min)) {\n min = h.created_at_i;\n }\n }\n }\n if (min === null) return null;\n return new Date(min * 1000).toISOString();\n } catch {\n return null;\n }\n}\n\n/** Run a list of async tasks with limited concurrency. */\nasync function mapLimit(\n items: T[],\n limit: number,\n fn: (item: T) => Promise,\n): Promise {\n const out: R[] = new Array(items.length);\n let next = 0;\n async function worker() {\n while (next < items.length) {\n const idx = next++;\n out[idx] = await fn(items[idx]);\n }\n }\n const workers = Array.from({ length: Math.min(limit, items.length) }, worker);\n await Promise.all(workers);\n return out;\n}\n\n/**\n * Refresh the cache: fetch the index, add any new essays (resolving their\n * dates), and backfill essays that still lack a date.\n */\nexport async function refresh(): Promise {\n const cache = await readCache();\n\n const html = await fetchIndexHtml();\n const list = parseIndex(html);\n\n // New essays plus known essays that are still missing a date.\n const needDate: { title: string; url: string; position: number }[] = [];\n for (let i = 0; i < list.length; i++) {\n const e = list[i];\n const existing = cache.essays[e.url];\n if (!existing || existing.date === null) {\n needDate.push({ title: e.title, url: e.url, position: i });\n }\n }\n\n const dated = await mapLimit(needDate, 8, async (e) => ({\n ...e,\n date: await hnDate(e.title),\n }));\n\n for (const e of dated) {\n cache.essays[e.url] = { title: e.title, date: e.date, position: e.position };\n }\n\n cache.lastCheck = new Date().toISOString();\n await writeCache(cache);\n}\n\n/** Timestamp (ms) of the last successful check, or 0. */\nexport async function lastChecked(): Promise {\n const c = await readCache();\n if (!c.lastCheck) return 0;\n const v = Date.parse(c.lastCheck);\n return Number.isNaN(v) ? 0 : v;\n}\n\nexport interface Essay {\n title: string;\n url: string;\n date: string | null;\n}\n\n/** All essays, newest first (by date, then list position). */\nexport async function fetchEssays(): Promise {\n const c = await readCache();\n const essays = Object.entries(c.essays).map(([url, e]) => ({\n url,\n title: e.title,\n date: e.date,\n position: e.position,\n }));\n essays.sort((a, b) => {\n const da = a.date ? Date.parse(a.date) : -Infinity;\n const db = b.date ? Date.parse(b.date) : -Infinity;\n if (db !== da) return db - da;\n return a.position - b.position;\n });\n return essays.map(({ url, title, date }) => ({ url, title, date }));\n}\n\nexport function esc(s: string): string {\n return s.replace(/&/g, \"&\").replace(//g, \">\")\n .replace(/\"/g, \""\");\n}\n\nexport function buildRss(essays: Essay[], selfUrl: string): string {\n const items = essays.map((e) => {\n const link = BASE + e.url;\n const dateEl = e.date ? `${new Date(e.date).toUTCString()}` : \"\";\n const desc = `${esc(e.title)} (Paul Graham).`;\n return `\\n ${esc(e.title)}\\n ${esc(link)}\\n ${esc(link)}\\n ${dateEl}\\n ${desc}\\n `;\n }).join(\"\\n \");\n\n return `\n\n \n Paul Graham’s Essays\n ${BASE}/articles.html\n Paul Graham’s essays, updated automatically from paulgraham.com.\n \n ${items}\n \n`;\n}\n","import { blob } from \"https://esm.town/v/std/blob/main.ts\";\nimport { load } from \"npm:cheerio@1.0.0\";\n\nconst INDEX_URL = \"https://www.berkshirehathaway.com/letters/letters.html\";\nconst CACHE_KEY = \"berkshire-shareholder-letters-feed\";\n\nexport interface FeedItem {\n title: string;\n url: string;\n date: string | null;\n summary: string | null;\n}\n\ninterface Cache {\n lastCheck: string | null;\n items: FeedItem[];\n}\n\nasync function readCache(): Promise {\n return (await blob.getJSON(CACHE_KEY)) ?? { lastCheck: null, items: [] };\n}\n\nasync function scrapeLetters(): Promise {\n const response = await fetch(INDEX_URL);\n if (!response.ok) throw new Error(`letters index: HTTP ${response.status}`);\n const $ = load(await response.text());\n const byYear = new Map();\n\n for (const element of $(\"a[href]\").toArray()) {\n const anchor = $(element);\n const href = anchor.attr(\"href\");\n const yearText = anchor.text().trim();\n if (!href || !/^\\d{4}$/.test(yearText)) continue;\n if (!/^\\d{4}(?:ltr)?\\.(?:html|pdf)$/i.test(href)) continue;\n const year = Number(yearText);\n if (year < 1965 || year > new Date().getUTCFullYear()) continue;\n const url = new URL(href, INDEX_URL).href;\n byYear.set(year, {\n title: `${year} Berkshire Hathaway shareholder letter`,\n url,\n date: null,\n summary: `Berkshire Hathaway annual letter for ${year}.`,\n });\n }\n\n const items = [...byYear.entries()].sort(([a], [b]) => b - a).map(([, item]) => item);\n if (items.length < 40) throw new Error(`index returned only ${items.length} letters`);\n return items;\n}\n\nexport async function refresh(): Promise {\n const items = await scrapeLetters();\n await blob.setJSON(CACHE_KEY, { lastCheck: new Date().toISOString(), items } satisfies Cache);\n}\n\nexport async function lastChecked(): Promise {\n const value = (await readCache()).lastCheck;\n const parsed = value ? Date.parse(value) : Number.NaN;\n return Number.isNaN(parsed) ? 0 : parsed;\n}\n\nexport async function fetchItems(): Promise {\n return (await readCache()).items;\n}\n\nfunction escapeXml(value: string): string {\n return value.replace(/&/g, \"&\").replace(//g, \">\")\n .replace(/\"/g, \""\").replace(/\\x27/g, \"'\");\n}\n\nexport function buildRss(items: FeedItem[], selfUrl: string): string {\n const body = items.map((item) => `\n ${escapeXml(item.title)}\n ${escapeXml(item.url)}\n ${escapeXml(item.url)}\n ${escapeXml(item.summary ?? item.title)}\n `).join(\"\\n \");\n\n return `\n\n \n Berkshire Hathaway shareholder letters\n ${INDEX_URL}\n Berkshire Hathaway shareholder letters from the official archive.\n en-us\n \n ${body}\n \n`;\n}","createdAt",["D",1788145959495],"branchVersion","branchId","acb0af00-730c-4ddc-ac11-2551e6da7020","id","666d7a52-e043-49ef-8d59-acb2fa08c994","berkshire-letters-rss",["D",1788145826110],"deletedAt","legacyEnvVars","privacy","public","httpPrivacy","description","RSS feed for Berkshire shareholder letters","imageUrl","41fc127f-5d67-4e9d-aaf5-7f1531438361","pinnedFileId","01a055cb-fa00-72a2-9988-e0c6c49eb841","pinPreview",true,"pinMigrated","screenshotVersion","likeCount",0,"orgId","9dd2e27d-d103-40dd-bb45-da8e837d341a","tursoName","666d7a52e04349ef8d59acb2fa08c994","tursoHostname","666d7a52e04349ef8d59acb2fa08c994-val-town.aws-us-east-2.turso.io","tursoMeasuredAt",["D",1789534872841],"pinned","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm","user",{"_103":102,"_110":106},"org",{"_103":102,"_104":-5,"_105":106,"_18":55,"_107":108,"_109":25},"branch",{"_13":89,"_32":90,"_43":42,"_91":44,"_92":93,"_94":95,"_96":50,"_97":98,"_99":65,"_100":45,"_101":102},"isOwned",8,"main","projectId","forkedBranchId","adc2f96c-f87b-444e-97b5-b08189b7e0e6","forkedBranchVersion",14,"forkedProjectPrivacy","forkedProjectId","af97ad78-f1f9-48ff-a00d-c401c8b9b233","forkedProjectOrgId","projectName","userHandle","ivryb","handle","displayName","avatar","https://img.clerk.com/eyJ0eXBlIjoicHJveHkiLCJzcmMiOiJodHRwczovL2ltYWdlcy5jbGVyay5kZXYvb2F1dGhfZ2l0aHViL2ltZ18yZkFtS1dWcU5NYWRjQkN5MGVCUTVRS0Z5SkEifQ","tier","free","canRestrictVals","profileImageUrl","branchName","","adoptOrgId",{"_43":44,"_32":45,"_38":212,"_47":-5,"_48":25,"_49":50,"_51":50,"_52":53,"_54":-5,"_18":55,"_56":57,"_58":59,"_60":-5,"_61":-5,"_62":63,"_64":65,"_66":67,"_68":69,"_70":213,"_72":-5,"_73":74,"_75":74,"_76":77,"_78":77,"_79":77,"_80":74,"_81":74,"_82":214,"_84":215,"_86":216,"_88":25},"tree",[191,192,193,194,195],"files",[191,192,193,194,195],"counts",{"_181":63,"_182":152,"_183":63,"_184":185,"_186":187},"contents",[143,144,145,146,147],"upstreamProject",{"_43":98,"_32":140,"_52":141,"_54":-5,"_49":50,"_82":142},"upstreamBranch",{"_43":93,"_38":135,"_136":137,"_32":90,"_91":98,"_13":95,"_92":-5,"_94":-5,"_138":139,"_47":-5,"_88":25,"_49":50,"_101":102,"_64":65,"_100":140},"emptyEnvVars","flags",[],"hasHttpFiles","grants","orgs","customDomains",{},["D",1787640981828],"updatedAt",["D",1787723249634],"lockfileKey","1787723250254-esm.town_v_ivryb_pg-essays-rss@14-main","pg-essays-rss","RSS feed of Paul Graham's essays",{"_103":102,"_110":106},{"_43":175,"_32":176,"_30":177,"_151":-5,"_41":42,"_40":152,"_153":25,"_38":178,"_13":63,"_33":20,"_35":179,"_24":25,"_156":-5,"_27":180},{"_43":168,"_32":169,"_30":57,"_151":-5,"_41":42,"_40":152,"_153":25,"_38":170,"_13":63,"_33":171,"_35":172,"_24":25,"_156":173,"_27":174},{"_43":161,"_32":162,"_30":163,"_151":-5,"_41":42,"_40":152,"_153":25,"_38":164,"_13":63,"_33":165,"_35":166,"_24":25,"_156":-5,"_27":167},{"_43":158,"_32":28,"_30":31,"_151":-5,"_41":42,"_40":14,"_153":25,"_38":159,"_13":152,"_33":34,"_35":37,"_24":25,"_156":-5,"_27":160},{"_43":148,"_32":149,"_30":150,"_151":-5,"_41":42,"_40":152,"_153":25,"_38":154,"_13":63,"_33":34,"_35":155,"_24":25,"_156":-5,"_27":157},"2a76dec3-7666-4aae-8258-bcd61fafa3a7","seed.ts","01a055cb-fa00-72a2-9988-eea57655884a","parentId",1,"latest",["D",1788145826110],"import { refresh, fetchEssays } from \"./scraper.ts\";\n\n// One-time backfill: fetch the full essay index and resolve publication dates\n// for every essay. Run via: run_file ivryb/pg-essays-rss seed.ts\nawait refresh();\nconst essays = await fetchEssays();\nconst withDate = essays.filter((e) => e.date).length;\nconsole.log(`Seeded ${essays.length} essays (${withDate} with a date).`);\n","label",[149],"22bab4f2-20a8-4f68-8a3c-af3209cd2fd2",["D",1788145959495],[28],"380154a0-88af-4036-8a50-6432e471ab5a","refresh.ts","01a055cb-fa00-72a2-9988-e42ac92f1cd3",["D",1788145826110],"interval","import { refresh } from \"./scraper.ts\";\n\n/** Keeps the cached essay index warm so feed requests stay fast. */\nexport default async function (): Promise {\n await refresh();\n console.log(\"pg-essays-rss refreshed\");\n}\n",[162],"0124f75d-aace-452d-b032-420022eb676e","main.ts",["D",1788145826110],"http","import { refresh, lastChecked, fetchEssays, buildRss } from \"./scraper.ts\";\n\n// Re-check the index at most hourly on request; the interval job keeps it warm\n// in between so responses are fast.\nconst STALE_MS = 60 * 60 * 1000;\n\nexport default async function handler(req: Request): Promise {\n try {\n const checked = await lastChecked();\n if (Date.now() - checked >= STALE_MS) {\n await refresh();\n }\n } catch (err) {\n // Don't fail the request if a refresh hiccups — serve stale data if we have it.\n console.error(\"refresh failed:\", err);\n }\n\n const essays = await fetchEssays();\n if (essays.length === 0) {\n return new Response(\"Feed not ready yet — try again in a minute.\", {\n status: 503,\n });\n }\n\n const xml = buildRss(essays, req.url);\n return new Response(xml, {\n status: 200,\n headers: { \"Content-Type\": \"application/rss+xml; charset=utf-8\" },\n });\n}\n","berkshire-letters",[169],"79a13ea2-a7e8-4dcd-bf9b-b40e4a7594f3","README.md","01a055cb-fa00-72a2-9988-de252711f305",["D",1788145826110],"# pg-essays-rss\n\nAn automatically-updated RSS feed of **[Paul Graham's essays](https://paulgraham.com/articles.html)**.\n\n**Feed URL:** \n\n## How it works\n\n- `scraper.ts` fetches `articles.html`, parses the ordered list of essays, and resolves each essay's publication date from the earliest `paulgraham.com` submission on the Hacker News Algolia API. The feed data is cached as a single JSON document in the val's **blob storage** (`std/blob`), which keeps it independent of the SQLite client.\n- `main.ts` (HTTP) serves an RSS 2.0 feed from that cache. When a request finds the cache older than an hour it re-checks the index (cheap) and only resolves dates for newly-seen essays, so responses stay fast.\n- `refresh.ts` (interval, cron `0 */3 * * *` UTC) keeps the index warm between requests.\n- `seed.ts` is a one-time backfill script — run it once to populate the database with all ~230 essays and their dates (`run_file ivryb/pg-essays-rss seed.ts`).\n\nNew essays are picked up the next time the index is re-checked (hourly on request, or every 3 hours via the interval), so the feed stays current without hitting Paul Graham's server on every read.\n\n## Notes\n\n- Publication dates are approximate (inferred from Hacker News submissions).\n- The archive contains ~230 items; some very old essays predating Hacker News may have no date.\n\n## Files\n\n| File | Type | Purpose |\n| --- | --- | --- |\n| `main.ts` | http | Serves the RSS feed at `/` |\n| `scraper.ts` | script | Index scraping, date resolution, caching |\n| `refresh.ts` | interval | Warm-up job (every 3 hours) |\n| `seed.ts` | script | One-time backfill |\n",[176],"pulls","branches","forks","envVars",{"_188":63,"_190":63},"connections",{"_188":63,"_189":63},"count","unfilledCount","emptyCount",{"_43":175,"_32":176,"_30":177,"_151":-5,"_41":42,"_40":152,"_153":25,"_38":209,"_13":63,"_33":20,"_35":179,"_24":25,"_156":-5,"_197":210,"_27":211},{"_43":168,"_32":169,"_30":57,"_151":-5,"_41":42,"_40":152,"_153":25,"_38":206,"_13":63,"_33":171,"_35":172,"_24":25,"_156":173,"_197":207,"_27":208},{"_43":161,"_32":162,"_30":163,"_151":-5,"_41":42,"_40":152,"_153":25,"_38":203,"_13":63,"_33":165,"_35":166,"_24":25,"_156":-5,"_197":204,"_27":205},{"_43":158,"_32":28,"_30":31,"_151":-5,"_41":42,"_40":14,"_153":25,"_38":200,"_13":152,"_33":34,"_35":37,"_24":25,"_156":-5,"_197":201,"_27":202},{"_43":148,"_32":149,"_30":150,"_151":-5,"_41":42,"_40":152,"_153":25,"_38":196,"_13":63,"_33":34,"_35":155,"_24":25,"_156":-5,"_197":198,"_27":199},["D",1788145826110],"children",[],[149],["D",1788145959495],[],[28],["D",1788145826110],[],[162],["D",1788145826110],[],[169],["D",1788145826110],[],[176],["D",1788145826110],["D",1789534872841],{"_103":102,"_110":106},{"_103":102,"_104":-5,"_105":106,"_18":55,"_107":108,"_109":25},{"_13":89,"_32":90,"_43":42,"_91":44,"_92":93,"_94":95,"_96":50,"_97":98,"_99":65,"_100":45,"_101":102},"ctx",{"_82":-5,"_132":250,"_251":252,"_253":254,"_255":25,"_256":25,"_257":25,"_258":59,"_259":-5,"_260":-7,"_261":-7,"_262":63,"_263":25,"_264":25,"_265":-5,"_266":25,"_267":25,"_268":25,"_269":-5},"ENV",{"_221":222,"_223":224,"_225":90,"_226":227,"_228":229,"_230":231,"_232":233,"_234":235,"_236":237,"_238":239,"_240":241,"_242":243,"_244":245,"_246":247,"_248":249},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","5015b8df46402d8bba4f485ecc37aec202b0e06b","RENDER_GIT_BRANCH","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","CLOUDFLARE_TURNSTILE_SITE_KEY","0x4AAAAAACtaIZJH7s9rBMZ6","GOOGLE_CLIENT_ID","875989696899-jtsqkhlbdv5jrlsk8631p7ipjg3c6o62.apps.googleusercontent.com","STRIPE_PUBLISHABLE_KEY","pk_live_51N2zvWEpZTTLipIkVEwXrUG9AT0yeOOEYxszarZT1jpsZV58Ly6wlg3SltZpHN26CUalOBmUSY7adQRQiSkCZz2500qSXNkf8Y","API_URL","https://api.val.town",[],"preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","preferencesCloseBrackets","preferencesChatPanelState","preferencesDprint","preferencesDenofmt","notificationCount","flagDenols","isBusinessOrgMember","businessOrgHandle","isBusinessOrgAdmin","isHigherTownieTier","navCollapsed","currentOrgId"]

Diff · v2