Skip to content

How the daily Hukamnama is verified

hukamnamaverification

The Hukamnama page shows the Hukamnama from Sri Harmandir Sahib, Amritsar, for each day. This article sets out how a day gets there, so you can judge for yourself how much to trust it. The short version: the text you read is never copied from anywhere. It is the same verified text as the Knowledge Base, located by an exact match, and a day that cannot be proven is not published.

Where the day comes from

The Shiromani Gurdwara Parbandhak Committee (SGPC) publishes the Hukamnama each morning on hs.sgpc.net, dated and with its Ang. Its main site, sgpc.net, shows the opening of the same Hukamnama separately. Our service reads both pages politely: every minute from 05:00 to 09:00 in Amritsar, then every ten minutes until 14:00. It never reads them more often than that.

Every date here is the Amritsar date. A page is used only when its own date stamp is today’s date in Amritsar.

How a day is located

SGPC’s page is used only to find where the Hukamnama is. Its words are compared with the lines of our text around the Ang it names, and the day resolves only when one run of whole lines matches every word, in order. There is no “close enough”. The comparison ignores only what is not a word:

  • spaces, the dandas (॥ and ।) and numerals: verse numbers, and the numeral in a label such as Mahalla 4 (a label is never matched on its own: the verses around it must match too);
  • invisible formatting characters;
  • one character that SGPC’s pages use for the subjoined ha (the sign U+0A51, where the printed Granth has a virama and ha), a known legacy of font conversion. Our own text never contains that character, so the mapping cannot join two different words.

The run must also end where a shabad ends (or, in a Vaar, where a pauri ends), so a page saved half-way is not taken as complete. The page must then stay the same for two minutes. If sgpc.net already shows today’s opening, it must agree with the match, on the Ang and on its opening words.

The heading printed above the shabad (its raag, the Guru and the invocation, Ik Onkar) comes from our text, never from SGPC’s page.

What is shown

What you read are our own lines, exactly as stored and cited by Ang. Transliteration is a separate reading aid, and so is the English translation when it is shown. Neither is part of the Gurmukhi, and both are labelled.

Each published day records two fingerprints: one for the whole text it was resolved in, and one for the exact lines of the day (a SHA-256 hash). The website reads the day’s lines again from its own copy of the text and checks them against that hash before it renders the page, and the feeds and the embed do the same. The public API’s files are written only after the same check. If the lines do not match, the day is not shown. The developers page explains how anyone using the API can make the same check.

When a day cannot be proven

If any of these steps fails, the day is held: nothing is published for it, and the page goes on showing the most recent Hukamnama under its own date, never as today’s. A person then compares SGPC’s page with the text by hand. If the difference is on SGPC’s side, the day is approved by hand for the exact lines, and each approval is recorded. We would rather be late than wrong.

Every held or corrected day is listed in the corrections log.

If something is wrong

If a published day turns out to be the wrong passage, or its lines are incomplete, it is corrected: the day keeps its first publication time, gains a “Corrected on” note on its page and in the API, and appears in the corrections log. To report an error, write to support. A person reads every report.