Charter
Compute faithfully. Cite everything. Never sell fear.
The classical tradition argued with itself: Varahamihira called some older yogas impossible, Kautilya warned kings against over-consulting the stars, and the Brihat Samhita condemned fake astrologers. This charter keeps that critical voice — as product rules, not small print.
Eight commitments
- Tendencies, not fate. A chart is presented as tendencies to work with. The texts themselves say so: the Mahabharata calls destiny the seed and effort the soil (Anushasana Parva 6); the Yoga Vasishtha calls fate a name for past action that present action overcomes; the Tirukkural (ch. 62) says effort earns its wage even against fate.
- No fear-based predictions. The engine can compute the classical longevity, maraka and widowhood rules, but they are never shown as personal predictions. Classical text on them appears only on clearly framed history pages.
- Nothing sold as a remedy. No gemstones, pujas, yantras or "parihara" packages, and no affiliate links to them. The Brihat Parashara Hora Shastra does not prescribe gemstones at all. Where a traditional practice is mentioned, it is free conduct, charity or devotion, and labelled "traditional, untested".
- Every statement has a source. Each rule names its text, edition, chapter and verse, and whether that edition is public domain. Variant readings are shown, not smoothed over. Legends are labelled as legends, history as history.
- Declared conventions. The ayanamsha, sunrise rule, node type and engine are printed next to the results, because those choices — not the mathematics — decide whether a panchanga is right to the minute.
- No dark patterns. No countdown banners, fake scarcity or fear-to-purchase flows (India's CCPA guidelines on misleading advertisements, 2022, and on dark patterns, 2023).
- Your birth data is yours. Charts are computed in your browser. Birth data is personal data under India's DPDP Act 2023; when accounts arrive they will be opt-in, encrypted, deletable in one click, and never used to train AI.
- AI only explains, and must cite. When an AI explainer arrives, it will never compute positions — the engine does — and every sentence it writes must point to a rule card. Uncited sentences are blocked.
Method: how this preview computes
- Positions — Ganita, our own engine: no third-party astronomy code. Planets from the VSOP87D theory (Bretagnon & Francou, 1988); the Moon from ELP2000-82B (Chapront-Touzé & Chapront, 1988), with its slow drift against NASA's current DE441 ephemeris measured and removed; nutation and sidereal time from the IAU 1980/1982 models; light-time, the FK5 frame correction and annual aberration applied. The theories' official data come from the Strasbourg astronomical data centre (CDS catalogues VI/81 and VI/79).
- Measured accuracy: against NASA JPL Horizons (DE441) at 80 instants from 1700 to 2150 that the engine was never fitted to, every body is within 1″ (one 3,600th of a degree); from 1900 to 2100 the Moon is within 0.3″ and the Sun within 0.2″. An arcsecond of the Moon is about two seconds of a tithi's end time.
- Time: clock times are converted with NASA JPL's tabulated TT − UTC from 1962 on (exact leap seconds); before 1962, with the Espenak–Meeus ΔT model, whose uncertainty grows into the past (about ±1 minute around 1000 CE), so historical charts carry a wider error band.
- Ayanamsha: Lahiri (Chhitrapaksha), India's official standard: 23°15′00.658″ mean at 21 March 1956 0h TT, carried forward by IAU 2006 general precession in longitude, plus nutation for the true value.
- Day and sunrise: the panchanga day runs from sunrise to sunrise. Sunrise is the Sun's upper limb crossing the horizon with standard atmospheric refraction. Other traditions (centre of the disc, no refraction) differ by two to three minutes.
- Limbs: tithi = (Moon − Sun) ÷ 12°; nakshatra = Moon ÷ 13°20′; yoga = (Sun + Moon) ÷ 13°20′; karana = half a tithi — as in Surya Siddhanta II.64–69 (Burgess 1860). End times are found by bisection to one second.
- Nodes: Rahu and Ketu are the mean lunar nodes (Meeus, Astronomical Algorithms, 47.7).
- Lagna: the ecliptic's rising point from apparent sidereal time and true obliquity, then made sidereal.
How we check it
The tests run on every build and are part of the repository:
- Positions against NASA JPL: 560 reference positions (7 bodies × 80 instants, 1700–2150) from JPL Horizons, compared in TT so only our theories are measured. A drift of a fraction of an arcsecond fails the build.
- Sankranti moments against Drik Panchang: Makara 2026 at 3:13 PM IST and Mesha 2026 at 9:38–9:39 AM IST. Ours land within a minute of both. An early version that read the 1956 constant as already "true" landed six and a half minutes early — the test caught it.
- Known days: Diwali 2025 (21 October) is Krishna Amavasya, a Tuesday; 1 January 2000 is a Saturday.
- Invariants across a year: two karanas per tithi, Ketu exactly opposite Rahu, every end time after the moment asked about, every day bounded by its sunrises, every tithi boundary exact to the second.
- Geometry: with Aries on the meridian at the equator, Cancer rises.
- The texts' own numbers: the Ashtakavarga totals 337 for any chart (Brihat Jataka); Vimshottari totals 120 years and every sub-period sums to its parent (BPHS 46, 51); natural friendships rebuilt from Satya's rule match the Brihat Jataka's table exactly.
Next on the list: Lahiri against the Indian Astronomical Ephemeris year by year, documented charts against Jagannatha Hora, and a full year of tithi end times at Ujjain, Delhi and Chennai against published panchangas. The pass rates will be published here.
Evidence: what science has found
We would rather you heard this from us. Controlled tests have not supported astrology's predictive claims:
- In Shawn Carlson's double-blind test, published in Nature in 1985, participants picked their own chart reading at about chance. A 2009 reanalysis by Suitbert Ertel reported a borderline result (p = .054), so the debate over that one study continues.
- Geoffrey Dean and Ivan Kelly (2003) reviewed more than 40 controlled studies and found astrologers performing at about chance. People born minutes apart ("time twins") showed none of the similarities astrology predicts.
So Astro treats Jyotisha as a living textual tradition, a cultural heritage and a tool for reflection — computed with care, cited to its sources, and never sold as certainty. That is both more honest and closer to what the old texts themselves argued.
Sources and copyright
The research behind this site read the scanned texts on the Internet Archive. Only about 30–40 of the 4,784 astrology items there are clean, out-of-copyright editions. Archive.org's licence tags are uploader claims, so we do not rely on them: editions are quoted at length only where public-domain status could be established (for example Chidambaram Aiyar's 1905 Brihat Jataka, Burgess's 1860 Surya Siddhanta, Shamasastry's 1915 Arthashastra). Later translations are summarised and cited, never copied. The full catalogue is in the library.
The rules we follow
- India, Drugs and Magic Remedies (Objectionable Advertisements) Act 1954: no talisman, mantra or kavach is ever presented as a cure.
- India, Consumer Protection Act 2019 and CCPA guidelines (2022, 2023): no misleading claims, no dark patterns, disclosed endorsements.
- India, Digital Personal Data Protection Act 2023 and Rules (2025): notice, consent and deletion designed in before the duties take effect.
- Elsewhere: horoscopes are presented as reflection and culture, never as guaranteed outcomes, consistent with UK BCAP and French ARPP advertising codes.