Haggas Nr. 145

[31 Jul 2026 inbox] THE LARGEST PARCEL IN THIS MODEL - 17,060 morgen, the loan farm Hagga's plus the annexed Government ground Callabash Kraal, surveyed Aug 1833 by Stretch for P. van der WesthuizenH and titled 15 Apr 1841 on Bf.W.Q.5-31 (DOC-0131). Named on the 1913 Witberg diagram as Witberg's north-eastern neighbour, which is why it is here. WHERE IT IS, established by Frans on the CSG viewer: the WILLOWMORE division, EASTERN CAPE - LPI prefix C0830000, e.g. portion 6/145 is C08300000000014500006 - lying directly EAST of Oorlogs Kloof (172, 175) and Vrolykheid (177), with the PROVINCIAL BOUNDARY running between. That is why farm 145 in our Prince Albert pull is Roode Punt Wes and in our Beaufort West pull is Truters Kuil: this project has never pulled Willowmore. NO GEOMETRY here for that reason, and deliberately not matched to a same-numbered farm in the wrong division - which is exactly the trap 9 warns about. Pulling the Willowmore division would map it and is the obvious next step. In 1833 this was all still 'Beaufort District', which was vast before the later divisions were cut, and the ward was TRAKA - the Willowmore-area Traka, not the Tarka near Tarkastad. CORRECTED SAME DAY: calling this 'the Willowmore-area Traka, not the Tarka near Tarkastad' was half wrong and misleading. TRAKA IS OUR OWN WARD. Vrolikheid (DOC-0027), Rondabel (DOC-0023), Oorlogs Kloof (DOC-0073) and Klipfontein (DOC-0112) each name 'Field Cornetcy of Traka' on their OWN 1833 diagrams, as do Klein Sleutelfontein, Oude Muure and Rhemhoogte 125. The 1935 Vrolykheid portion diagrams show Traka struck through and corrected by hand to KREDOUW, so the ward was simply renamed. Haggas sitting in the modern Willowmore division means the 1830s ward straddled ground that later fell either side of a provincial boundary - it does not make Traka an eastern place. See LEAD-0014, which this answers.

[31 Jul 2026] GEOMETRY WITHDRAWN - the pin was 104 km wrong. Frans, looking at the map: "wtf - haggas should next to oorlogskloof", and he is right: his CSG screenshot gives HAGGAS 145 as LPI C08300000000014500006, registration division C0830000 WILLOWMORE, province EASTERN, sitting against RE/172, 2/172 and RE/175 - which are our Oorlogs Kloof No. 172 and No. 175.

WHAT WENT WRONG, and it is the 9 trap in a new place: FARM NUMBERS REPEAT ACROSS REGISTRATION DIVISIONS. match_cadastre.py had only Prince Albert and Beaufort West polygons to match a "145" against, so it matched a DIFFERENT farm 145 and put Haggas at -33.231944,21.695017. The match report said so at the time - stated extent 17,060 mg = 14,612 ha against a cadastre polygon of 8,845 ha, a ratio of 0.518 - and the warning was printed and not acted on.

NOT REFETCHABLE HERE: Willowmore is EASTERN CAPE and data/fetch_cadastre.py talks to the WESTERN Cape Surveyor-General. Adding WILLOWMORE to that fetch returns zero features, which is the correct answer from the wrong service. Haggas now has NO polygon and is listed as unmapped, per 5: an unresolved place gets no pin, because a wrong pin is worse than a missing one. [TO VERIFY] pull from the Eastern Cape service, or georeference from the CSG viewer, keying on LPI C0830000/145.

[1 Aug 2026] *** WHERE HAGGAS ACTUALLY IS, from a document rather than a cadastre lookup. *** BOTH Oorlogskloof diagrams name Haggis on its boundary: Stretch's 1833 sheet has it on the SOUTH, and Melvill's 1863 resurvey has it on BOTH THE EAST AND THE SOUTH. Oorlogskloof's western boundary is Vrolykheid (PAR-000014). So Haggas sits east and south of Oorlogskloof, which sits east of Vrolikheid - all of it in the Traka field cornetcy, Prince Albert division. THAT IS THE DOCUMENTARY VERSION OF WHAT The geometry stays withdrawn (a wrong pin is worse than none, 5), but the search area is now a documented one: whatever polygon is right must touch Oorlogskloof No. 172 on its east and south.

[1 Aug 2026] *** LOCATED, FROM THE OFFICIAL CSG VIEWER. *** Frans opened it on the parcel itself and the panel reads: Farm Name HAGGAS 145 · 21 Digit Code C08300000000014500006 · Province EASTERN · Major Region WILLOWMORE · Major Code C0830000 · Parcel 145 Portion 6 · Geometry Area 13,966,629.69 m2 (1,396.7 ha) · Parcel Type FP. It is highlighted immediately EAST of RE/175 and 4/145, exactly where Stretch's 1833 sheet and Melvill's 1863 resurvey both put Haggis on Oorlogskloof's boundary. *** AND IT EXPLAINS THE WHOLE CONFUSION: BOTH THINGS WERE TRUE AT ONCE. *** Haggas really IS registered in WILLOWMORE (C0830000, Eastern Cape) - which is why matching it in the Prince Albert division found a different farm 145 and pinned it 104 km away - AND it really IS next to Oorlogskloof, exactly as The registration-division boundary runs right there. The mistake was assuming that Willowmore-registered meant far from Prince Albert. STILL NO GEOMETRY HERE, and the reason has changed: it is no longer that we do not know where it is, it is that we cannot DRAW it. The cadastre fetcher asks the WESTERN CAPE Surveyor-General's service, which does not serve an Eastern Cape division - geo/willowmore_farms.geojson came back with ZERO FEATURES and was written to disk as a 127-byte file, so the failure looked like a success. See LEAD-0053. The LPI above is now recorded and is the thing to fetch with.

*** RESOLVED 1 Aug 2026, AND IT WAS NEVER A SURVEY PROBLEM. *** Haggas is drawn. The polygon is E107C083000000000145000000, 14,216.4 ha, from the NATIONAL cadastre published by the Council for Geoscience, which carries the Eastern Cape and therefore Willowmore. Two independent confirmations: our 1833 sheet says 17,060 morgen = 14,612 ha, a 2.7% agreement; and portion 6 comes back at 1,396.7 ha, the exact figure on Frans's CSG screenshot. Its centroid is 16.1 km EAST of Oorlogs Kloof, which is what THE REAL CAUSE OF THE MONTH-LONG DELAY was not the Eastern Cape at all. fetch_cadastre.py wrote to data/geo/ and match_cadastre.py and build_site.py both read <repo>/geo/. The two directories had drifted, so expanding the pull from 3 divisions to 9 and adding the national Willowmore fetch changed nothing downstream. The symptom to remember: a successful fetch that does not move the match count.

[4 Aug 2026, SRC-249 + SRC-253] TWO PEOPLE DIED ON THIS FARM, and identifying it took two documents read seven months apart. Cornelis Petrus Janse van VuurenL died here on 4 September 1863 - his notice writes the place as "[?Hoeagus]", which was unreadable and matched nothing. His widow MARIA MAGDALENA ELIZABETH VAN DER WESTHUIZEN died here on 24 June 1880, and her notice writes it plainly: "at the farm Hagas District Prince Albert". A third spelling, Haagus, appears on FamilySearch 9ZPK-FYB. One word, three hands. THE STRONGEST PART OF THE IDENTIFICATION IS THE SURNAME: this farm was surveyed in August 1833 for P. VAN DER WESTHUIZEN, and the woman who died on it was a van der Westhuizen. Recorded as RESIDENCE ONLY - the notices say where they lived and died, not what they owned, and no deed here puts either of them on the title. The district discrepancy (Willowmore in the cadastre, Prince Albert on both notices) is recorded unresolved.

[SRC-321, 7 Aug 2026] A FARM CEMETERY IS RECORDED ON THIS GROUND BY eGGSA: 'Gamkaspoortdam, Witpoort 145', Cemetery ID 8155, at -33.255133, 21.680567. Photographs and captions at https://graveswcape.eggsa.org/PRINCE-ALBERT-Rural/Witpoort-145/ - linked, not copied.