Simutrans TimetablesCZ · fork 122.0 · pak128 VZ
mirror/pipeline/brno/method-check-2026-10-25.mdView on GitHub ↗

City sheets: row-by-row method check (25 Oct 2026)

Sheets: Brno Tram Fleet from 29 Sep (brno, state from-2026-09-29), Brno Bus Peak Fleet (brno-bus), Brno Trolleybus Peak Fleet (brno-trolleybus), Plzeň Tram Peak Fleet (plzen, 2026-09-27), Plzeň Trolleybus Peak Fleet (plzen-trolleybus), Plzeň Bus Peak Fleet (plzen-bus, scripts of plzen-trolleybus with --base), Pardubice MHD Peak Sheet (pardubice). Built on top of today's offset-origin fix (pipeline/audit/city-changes-2026-10-25.md), which is unchanged.

Owner's request: "verify all sheets with this level of detail … that's the point of you doing it, so you'll not produce incorrect results".

1. What was checked, and how

Every Simutrans row (Brno tram 26, Brno trolleybus 36, Brno bus 90, Plzeň tram 8, Plzeň trolleybus 26, Plzeň bus 47, Pardubice 69) was re-read against the real departures of its entry for the whole afternoon (13:00–20:00), taken from the same GTFS / JDF data and filters the pipeline uses (depot runs out, minute subsets and aliases as in the state, Pardubice merged minutes as in network.json). Per row:

  • Typed values: every value = (real − 60 − 1) mod every at turns/loops (all entries on these sheets except Brno tram Mifkova and Nemocnice Bohunice, which keep real − 60). Checked from each value, never by zipping lists.
  • Exact stretch: the longest contiguous stretch where the pattern's slots and the real departures coincide minute for minute; flagged when it has fewer than 2 departures, or when a multi-slot pattern is exact for less than one interval (a "one-cycle" fit: any 8 departures inside 48 min are trivially "exact" every 48 with 8 slots).
  • Peak: the line's own busiest 2 h in 13:00–18:00, or 15:00–17:00 when that is within one departure of it (kb city-fleet-sheets); departures of that window outside the exact stretch were listed.
  • Better pattern: every interval dividing 1440 up to 60 (and 72–120), ≤ 8 slots, at the line's rate and at every other rate, ranked by how much of the peak window its exact stretch covers; the pattern's bus need (follow a bus round the slots, real running times, 3-min layover) compared with the sheet's need.
  • Same rate at both ends of every Simutrans line (all legs of a circuit), folding, ≤ 8 slots, offset < interval.
  • Approximate rows: largest deviation, and whether an exact or simpler pattern was missed.

Rule for changing a row (owner's rules; the sheets are already in the game, so as few moves as possible): change it when it breaks a hard rule (rate, < 2 departures, one-cycle fit, missed exact pattern behind an approximate label) and a better pattern exists that needs the same vehicles. Where the better pattern costs vehicles, the row is kept and the option is listed for the owner (section 5). No vehicle was moved: every rebuilt sheet's fleet.json has the same content as before.

2. Findings by sheet

Brno Tram Fleet from 29 Sep — no change

  • All 26 rows exact, typed values right, both ends at one rate.
  • 1, 4, 6, 9 run every 5 until about 16:30 (exact for 1.5–2 h ending 16:20–16:40), then every 6–7: the departures "outside the run" are the after-peak form. 5/5S, 7, 10, 2, 3 exact for 1.5–5 h.
  • 8 Bohunice → Mifkova: exact 14:47–16:37 except the 15:50 short trip that runs through to Mifkova (state skip, said on the sheet); 8S stops at 16:30 (peak reinforcement).
  • 12 every 14 (3 slots, 5-5-4 cycle): exact 14:41–16:24 and 14:55–16:29; values from the 01:00 origin (today's fix) verified.

Brno Trolleybus Peak Fleet — 1 entry changed

  • All 36 rows exact over 1–5 h, one rate per line (circuits 25 + 26 and 35 + 38 + 39 included).
  • 37N Mendlovo náměstí → Nemocnice Bohunice (bunching at a shared terminus): the row was the branch's irregular pre-16:33 minutes (every 60, :07 :20 :33 :53, exact 15:20–16:33) while 37J is in its post-16:41 form (every 15 :11), so in the game the two branches left Mendlovo náměstí 3–4 min apart (:07/:11, :53/:56) with a 12-min gap. After 16:33 37 alternates N :03 :18 :33 :48 / J :11 :26 :41 :56. Changed to every 15 :03 (exact 16:48–18:03), the same rate as its Bohunice end (every 15, exact 14:35–17:05). Need 4 (unchanged).
  • 33 Slatina → Hlavní nádraží: real :05/:18/:25/:38/:45/:58 is exact every 20 [5, 18]; the :12 slot is the documented designed return (designed_slots, 9 out against 6 back).
  • 34, 36, 37J change frequency inside the peak (README): their rows are the post-16:00 form (34 exact 16:00–18:00, 36 15:59–17:40, 37J 15:53–18:11 / 16:41–18:11). Their earlier forms are irregular (5 departures an hour with gaps; 34 every 10 with 15:08 and 15:20 missing). 34's every-10 form would need the same 4 trolleybuses but is exact only 15:18–16:18 / 15:30–16:00 — listed in section 5.

Plzeň Tram Peak Fleet — no change

  • 1 and 2 every 5 in their 5-min peak (1 exact 13:41–16:21; 2 exact one hour each end, the minute drifts by 1 at 15:16/15:22 and 15:43/15:49 — one drifting departure, same pattern).
  • 4 (one line, Košutka twice, every 8): exact 16:04–17:44 from Košutka, 16:48–17:50 from Bory/Univerzita. Before 16:00 (and at the far ends until about 16:45) the two branches run an irregular 3/6/9 pattern; no repeating pattern (any interval ≤ 60, ≤ 8 slots) is exact there. Documented state decision (core 16:00–17:30); kept.

Brno Bus Peak Fleet — 3 lines changed

Line Finding Change
48 Úzká – Prace, točna Úzká row approximate (every 60, :43 "not modelled"). The line runs every 30 at both core points all afternoon: Úzká :13/:43 13:13–18:13, Prace, točna :04/:34 13:34–18:34; every other hour the :43 / :34 trip starts or ends two stops further at Prace, Mohyla míru (GTFS: leaves Prace, točna 14:34, 16:34, 18:34; Úzká 15:43, 17:43 runs on). Both ends every 30, exact. New override field passing counts those departures for the match. Need 3 (unchanged; period: 60 lets the cycle close past the 17:43 extension trip, otherwise the 30-min wait guard drops the need).
72L Poliklinika Lesná – Technologický park Rate 3 voted from the 16:00 hour (Technologický park 3 departures because 16:48 runs to Královo Pole). The Lesná section runs 4 an hour all afternoon: Lesná :15/:25/:45/:55 exact 13:45–16:55 (14 departures), Technologický park :08/:18/:38/:48 exact 14:08–16:38 (13). Rate 4, every 30 [15, 25] / [8, 18]. Need 4 (unchanged). The 16:48 slot is the bus that goes to Královo Pole in reality (note on the sheet). With 72S (:28/:58) Technologický park now runs every 10.
75 Tomkovo náměstí – Slatina, nádraží Slatina row approximate (16:14 = tail of the every-15 form). Its peak form is every 15: Tomkovo :01 exact 14:16–15:46, Slatina :14 exact 14:29–16:29; it needs 4 buses, the sheet's need (3 + 1 on the road). Both ends every 15. Need 4 (unchanged).

Checked and kept: 47, 49P, 77L run every 15 / every 30 in 13:30–16:00 and thin out at 16:00 — their peak form costs one bus each (section 5). 77L Letiště stays approximate (Letiště 2 an hour against Úzká 1 from 16:08). 77A, E50, E75 (one-way peak lines) exact 2–4 h. All other rows: no exact pattern at the same rate covers more of the line's peak.

Plzeň Trolleybus Peak Fleet — 2 lines changed

Line Finding Change
12 Nová Hospoda – Božkov (8 an hour) Rows were every 60 with 8 slots and every 45 with 6 slots: "exact" only for one cycle (16:30–17:22, 15:59–16:44), ±4 min. The real headway thins steadily from 5 to 8 min, so nothing repeats exactly. Every 15 with two slots 7–8 min apart: Nová Hospoda :00 :08, Božkov :00 :07; ±2 min in the 16:00 hour, ±3 over 15:30–17:00 (better than before, 2 values instead of 8). Need 10 (unchanged).
16 Sídliště Bory → Doubravka (10 an hour) Every 48 with 8 slots, exact for one 48-min cycle only (the row type the owner rejected on Ostrava: "why is this not just…"). Every 6, real :00 mod 6 (±2 in the 16:00 hour, ±3 over 15:30–17:00), the same form as the Doubravka end. Need 12 (unchanged).

Checked and kept: 10 (Tylova every 12 combined, the two branches alternate exactly 15:20–16:32; the far ends shift by half an interval at 16:00 — approximate ±12, as labelled); 11, 14, 18, 49 exact; 13 Na Dlouhých exact 14:05–16:29 (16:41 slot is the short trip to Černice); 15 Borská pole drops the :07 bus that continues as 15N (correct); 15N, 22 Koterov irregular Škoda-shift trips (approximate, labelled). 15 at 5 an hour (+1 trolleybus) and 14 at 2 an hour (+2) would cost vehicles (section 5).

Plzeň Bus Peak Fleet — 6 lines changed

Line Finding Change
26 Bory – Lhota RATE: Bory 1 an hour (:30) against Lhota 2 (:04 :54). Real: 2 an hour Bory :20/:40, Lhota :04/:44 until the 16:04 return, then hourly; the 16:30 bus turns at Lhota for 16:54 (and 17:40 → 18:04). Both ends hourly: Lhota :54 (the 16:30 bus's turn). 16:04 not modelled. Need 1 (unchanged). Option: half-hourly form (+1 bus), section 5.
25 Bory – Bolevec Bolevec :15 was a single departure (the return of 15:37, earlier form). Bolevec :00, the turn of the hourly Bory :07 trips (16:07 → 17:00, 17:07 → 18:00); exact 17:00–18:00. Need 3.
29 Borská pole – Doubravka Rate 3 from the 16:00 hour; Doubravka approximate ±14. The line runs every 15 at both ends (Borská pole :04 exact 14:49–16:19, Doubravka :04 exact 15:49–16:49, 15:03–15:33 one min early). Its pattern needs 6 buses = the sheet's need. Both ends every 15 :04. Need 6.
33, 40 (33 + 40 loop, both directions) Rows were the 16:xx minutes as every 60 with 3–4 uneven slots, exact for under one hour. The loop runs every 15 both ways until about 16:30: 33 Hlavní nádraží :11 (14:26–16:26), 40 Sídliště Košutka :10 (15:10–16:40), 40 Hlavní nádraží :04 (14:34–16:34), 33 Sídliště Košutka :02 (15:02–16:32). Pattern need 4 each = the sheet's need. All four entries every 15 (33 rate 4). Need 4 + 4. Hlavní nádraží and Košutka now alternate 33/40 every 7–8 min, as in reality.
38 Doubravka – Červený Hrádek Červený Hrádek :08 was a single departure. A 6-min run: one bus does every 30 (or every 15). Every 30 both ends (Doubravka :00, Červený Hrádek :08, exact 15:00–16:08). Need 1.

Checked and kept: 21 Litice (every 60 :05/:20/:40 is the 16:00 hour; every 20 :00 would be exact 16:20–18:00 but matches fewer peak departures — marginal, left); 24 Bolevec (irregular, approximate); 32, 41, 41K approximate (irregular; no exact pattern at their rate); 51, 31, 35, 39, 20, 27, 28, 30, 26N exact; 42 consistent with 26 (Bory :00 26N, :15/:45 42, :30 26 = every 15, as after 16:00). Many rows reproduce only the 16:00 hour because these lines thin out at 16:00 (21, 24, 25, 26/42, 30P, 34, 41K): section 5.

A rebuild of either Plzeň sheet also rewrites one sentence in the pak table: it lists the local pak checkouts it read, and the -mhd and -pidops checkouts no longer exist on this machine. No object status changes (all still "requested").

Pardubice MHD Peak Sheet — 17 lines changed (fit_overrides, new)

city_systemic does not read Pardubice; checked here from network.json. Before: 16 Simutrans lines broke "both ends at one rate" — 2 (5 vs 3 an hour), 2P (1 vs 3), 3 (4 vs 3), 6 (4 vs 2), 9 (2 vs 1), 9S (1 vs 2), 10 (2 vs 1), 11 (3 vs 2), 12M (2 vs 1), 13 (6 vs 5), 14 (1.5 vs 2), 17 (1 vs 2), 30 (3 vs 2), and the daily 23, 24, 25 — and many rows were the 16:xx minutes repeated hourly (every 60 with up to 6 slots for every-10 lines), exact for under an hour, some for a single departure. Causes: the 16:00-hour fallback, minutes of other patterns merged into one end only, and no pair step. After: every non-daily line at one rate (checked), every typed value (real − 61) mod every, no fold, ≤ 8 slots. Vehicles and needs unchanged (cycle_need_peak sizes from the real trips). Every changed line was also sized from its new pattern (follow a bus round the slots, real running times, the stop's own turnaround up to 3 min): none needs more vehicles than the sheet gives it; 2, 3, 5, 10, 11 and 30 would need fewer (the sheet keeps the real-trip numbers).

  • Exact patterns missed: 13 every 10 both ends (exact 14:07–16:47 / 14:08–16:38; the hour fit failed on 16:59 vs 16:57), 5 Dubina every 10 (14:01–16:41), 1 Jesničánky :06/:30/:54 (14:30–16:30, was :58), 8 Dubinská every 30 :11 (14:41–16:11).
  • Merged minutes on one end only: 3 Hlavní nádraží :12/:27/:42 (own trips, 3 an hour like the return; the U Mlýnů :57 trips not modelled), 10 Hlavní nádraží hourly :22 (Nemošice/park :52 not modelled), 11 Dubina every 30 :15 (Rybitví trips and 16:35 not), 30 Hlavní nádraží every 30 :25 (3 Ohrazenice trips not).
  • Transition hour (the 16:00 hour is where the peak form ends): 2 + 2P in their peak form, exact 14:33–16:24 — 2 every 30 (Polabiny :13, Zámeček :12), 2P 4 an hour (Polabiny :03 :23 :33 :53, Pardubičky :04 :24 :34 :54); together every 10 from Polabiny as in reality. 6 + 6H every 20 (6 Vápenka :00, Rosice :16; 6H Vápenka :10; exact 14:10–16:20), together every 10 from Vápenka. 6H Hlavní nádraží is set at :18, 2 min after the real :16 :36 :56: at the real minute the pattern needs 4 buses, at :18 the bus from Vápenka (arriving :35) leaves at once and 3 do it, the sheet's need.
  • Irregular lines, one even pattern at the rate the need allows, set at a real bus turn, labelled approximate with the largest deviation (15:00–18:00): 9 hourly Hlavní nádraží :23 / Hůrka :48 (±29; every 30 would need a 2nd bus), 9S hourly :01 / Spojil :22 (±22), 12M hourly :51 / Mnětice :22 (±22; :51 is 1 min after the real 15:50 so one bus closes the 60-min round), 14 every 30 Polabiny :16 / Staré Čívice :10 (±7; slower option every 40, ±19), 17 every 30 :15 / Srch :13 (±13; within 3 min 15:13–16:46).
  • Kept: 4, 7, 18, 33, 30C, 1D, 21, 27, 20 (exact); 12, 15, 28, 8S, 88 (16:00-hour rows, rates equal, irregular timetables).
  • The sheet's method text said "Offset counts from midnight"; corrected to 01:00 (also on both Plzeň sheets).
  • Bunching scan at shared termini (all sheets): every pair of lines the model sends from one stop within 2 min also leaves that close in the real 15:00–17:00 timetable, except Pardubice Hlavní nádraží 17 (:45, approximate) and 15 (:45) — different routes, so left.

3. The Pardubice question for the owner (rows NOT changed)

Lines 16, 22, 23, 24, 25 (and the airport line 90) are every-1440 rows with each departure of the day as Offset + Also at. That breaks "no 1440 on bus sheets", and 23, 24, 25 also fail "both ends at one rate" as typed (5 vs 4, 8 vs 6, 3 vs 8 departures a day — buses pile up at one end). Pardubice is a city network, but these are its thin outer and shift lines. Which method should Pardubice use?

Option A — small-town method: bend onto a repeating ≤ 480 pattern Option B — big-city method: list, don't model
16 Hlavní nádraží – Dříteč (6 + 6 trips) every 480, the afternoon trips exact: Hlavní nádraží 12:00 14:04 17:10 19:25 (Offset 9, Also at 144, 179, 303), Dříteč 12:30 14:37 17:40 19:53 (Offset 39, Also at 172, 209, 336); 12 a day each way, the morning 5:55/8:29 bent to 6:04/9:10 (+9/+41). 1 bus (116). not modelled: −1 bus (116); its 8 stops lose service (Staré Hradiště, Psinek; Staré Hradiště, Brozany; Ráby, křižovatka; Ráby, prodejna; Ráby, Kunětická hora; Němčice, chaloupky; Němčice, rybníček; Dříteč)
22 Závodu míru – Nemošice runs hourly 13:24–16:50: every 60, Závodu míru :24 (Offset 23), Nemošice :50 (Offset 49), exact 4 trips each way; the 6:11/7:10 morning trips not modelled. 1 bus (113). Recommended under either option. same as A (an ordinary hourly afternoon line)
23 Staré Čívice – Hlavní nádraží (shift line) every 480: Staré Čívice 14:14 17:20 (Offset 19, Also at 313), Hlavní nádraží 13:28 17:55 (Offset 54, Also at 267); 6 a day each way, mornings bent (5:30 +44, 5:55 −27). 1 bus (67). The 12 Polabiny – Staré Čívice trips are not on it either way. not modelled: −1 bus (67); no stop served only by 23
24 Čepí, prodejna – Hlavní nádraží every 480: Čepí 14:57 15:47 16:53 (Offset 356, Also at 406, 472), Hlavní nádraží 14:30 15:20 16:25 (Offset 329, Also at 379, 444); 9 a day each way, early/late trips bent up to 2 h (4:56). 1 bus (68). The Starý Mateřov trips are not on it either way. not modelled: −1 bus (68); 3 stops lose service (Čepí, prodejna; Starý Mateřov, točna; Starý Mateřov, náves)
25 Dubina, sever – Staré Čívice (shift line, 8 out / 3 back) every 480: Staré Čívice 14:15 15:38 (Offset 314, Also at 397), Dubina 13:00 13:35 (Offset 239, Also at 274); 6 a day each way, the 5 morning shift trips bent to 5:00/5:35 (up to 95 min). 1 bus (188). not modelled: −1 bus (188); no stop served only by 25
90 airport (flight times) every 1440 was asked for by the owner (flight-bound); no ≤ 480 pattern fits 8:50/9:20 and 19:20/19:50. Keep as the one exception? same

Option A keeps every stop and bus but runs about twice the real trips a day on 16, 23, 24, 25 (each bus shuttles all day in the game). Option B is the rule for big cities (Ostrava, Prague, Brno, Plzeň): 4 Simutrans lines and 4 buses fewer, and 11 village stops (16, 24) without service on the sheet.

4. In-game changes (typed values; real minutes = value + 61 mod every, all entries are turns)

Brno trolleybus

  • 37N Mendlovo náměstí → Nemocnice Bohunice: every 60, Offset 6, Also at 19, 32, 52 → every 15, Offset 2 (real :03 :18 :33 :48).

Brno bus

  • 48 Úzká → Prace, točna: every 60, Offset 12 → every 30, Offset 12 (real :13 :43). Prace, točna → Úzká: every 60, Offset 3 → every 30, Offset 3 (real :04 :34).
  • 72L Poliklinika Lesná: every 60, Offset 24, Also at 44, 54 → every 30, Offset 14, Also at 24 (real :15 :25 :45 :55). Technologický park → Lesná: every 60, Offset 7, Also at 17, 37 → every 30, Offset 7, Also at 17 (real :08 :18 :38 :48).
  • 75 Tomkovo náměstí: every 30, Offset 15 → every 15, Offset 0 (real :01 :16 :31 :46). Slatina, nádraží: every 30, Offset 28 → every 15, Offset 13 (real :14 :29 :44 :59).

Plzeň trolleybus

  • 12 Nová Hospoda: every 60, Offset 5, Also at 13, 21, 29, 36, 43, 50, 57 → every 15, Offset 14, Also at 7 (real :00 :08 :15 :23 …). Božkov: every 45, Offset 35, Also at 43, 5, 12, 19, 27 → every 15, Offset 14, Also at 6 (real :00 :07 :15 :22 …).
  • 16 Sídliště Bory → Doubravka: every 48, Offset 41, Also at 0, 5, 11, 16, 22, 28, 34 → every 6, Offset 5 (real :00 :06 :12 …).

Plzeň bus

  • 25 Bolevec → Bory: every 60, Offset 14 → Offset 59 (real :00).
  • 26 Lhota → Bory: every 60, Offset 3, Also at 53 → Offset 53, no Also at (real :54).
  • 29 Borská pole: every 60, Offset 3, Also at 18, 32 → every 15, Offset 3 (real :04 :19 :34 :49). Doubravka: every 60, Offset 2, Also at 17, 47 → every 15, Offset 3 (real :04 …).
  • 33 Hlavní nádraží: every 60, Offset 10, Also at 25, 43 → every 15, Offset 10 (real :11 :26 :41 :56); Sídliště Košutka (as 40): every 60, Offset 9, Also at 24, 39 → every 15, Offset 9 (real :10 :25 :40 :55).
  • 40 Hlavní nádraží: every 60, Offset 3, Also at 18, 33, 53 → every 15, Offset 3 (real :04 :19 :34 :49); Sídliště Košutka (as 33): every 60, Offset 1, Also at 16, 31, 49 → every 15, Offset 1 (real :02 :17 :32 :47).
  • 38 Doubravka: every 60, Offset 59 → every 30, Offset 29 (real :00 :30). Červený Hrádek: every 60, Offset 7 → every 30, Offset 7 (real :08 :38).

Pardubice

  • 1 Jesničánky → Slovany: every 60, Offset 5, Also at 29, 57 → Also at 29, 53 (real :06 :30 :54).
  • 2 Polabiny → Zámeček: every 60, Offset 2, Also at 22, 32, 42, 57 → every 30, Offset 12 (real :13 :43). Zámeček: every 60, Offset 11, Also at 31, 56 → every 30, Offset 11 (real :12 :42).
  • 2P Polabiny → Pardubičky: every 60, Offset 12 → every 30, Offset 2, Also at 22 (real :03 :23 :33 :53). Pardubičky: every 60, Offset 3, Also at 23, 43 → every 30, Offset 3, Also at 23 (real :04 :24 :34 :54).
  • 3 Hlavní nádraží → Lázně Bohdaneč: every 15, Offset 11 → every 60, Offset 11, Also at 26, 41 (real :12 :27 :42).
  • 5 Dubina → Dukla: every 60, Offset 0, Also at 10, 20, 30, 40, 51 → every 10, Offset 0 (real :01 :11 …).
  • 6 Vápenka → Rosice: every 60, Offset 19, Also at 29, 56, 59 → every 20, Offset 19 (real :00 :20 :40). Rosice: every 60, Offset 15, Also at 42 → every 20, Offset 15 (real :16 :36 :56).
  • 6H Vápenka → Hlavní nádraží: every 60, Offset 9, Also at 41 → every 20, Offset 9 (real :10 :30 :50). Hlavní nádraží → Vápenka: every 60, Offset 15, Also at 37 → every 20, Offset 17 (real :18 :38 :58).
  • 8 Dubinská → Hlavní nádraží: every 60, Offset 10, Also at 52 → every 30, Offset 10 (real :11 :41).
  • 9 Hlavní nádraží → Hůrka: every 60, Offset 22, Also at 51 → Offset 22, no Also at (real :23).
  • 9S Spojil → Hlavní nádraží: every 60, Offset 0, Also at 21 → Offset 21, no Also at (real :22).
  • 10 Hlavní nádraží → Ostřešany: every 30, Offset 21 → every 60, Offset 21 (real :22).
  • 11 Dubina → Globus: every 60, Offset 14, Also at 34, 56 → every 30, Offset 14 (real :15 :45).
  • 12M Hlavní nádraží → Mnětice: every 30, Offset 2 → every 60, Offset 50 (real :51).
  • 13 Dubina → Ohrazenice: every 60, Offset 6, Also at 16, 26, 36, 46, 58 → every 10, Offset 6 (real :07 :17 …). Ohrazenice: every 60, Offset 7, Also at 17, 27, 37, 48 → every 10, Offset 7 (real :08 :18 …).
  • 14 Polabiny → Staré Čívice: every 40, Offset 0 → every 30, Offset 15 (real :16 :46). Staré Čívice → Polabiny: every 60, Offset 10, Also at 35 → every 30, Offset 9 (real :10 :40).
  • 17 Hlavní nádraží → Srch: every 60, Offset 15 → every 30, Offset 14 (real :15 :45). Srch: every 60, Offset 12, Also at 45 → every 30, Offset 12 (real :13 :43).
  • 30 Hlavní nádraží → Univerzita: every 60, Offset 9, Also at 24, 54 → every 30, Offset 24 (real :25 :55).

Vehicles: none in, none out, on any sheet (fleet.json unchanged; needs unchanged — only 48's and 38's cycle minutes changed). Every changed line on every sheet was sized from its new pattern as well: none needs more than the sheet assigns.

5. Lines modelled in their wind-down (owner's choice; not changed)

These lines run their peak earlier than 16:00 and thin out in the 16:00 hour the Brno bus and Plzeň pipelines fit to. Their peak form is exact for 1–3 h but needs more buses than the sheet has (pattern sizing, 3-min layover), so it was not applied:

Line Peak form (exact) Buses
Brno bus 47 every 15 (Hlavní nádraží :04 14:19–15:34) +1
Brno bus 49P every 15 (Mariánské náměstí :05 14:35–15:50) +1
Brno bus 77L every 30 (Úzká :08 13:08–16:08, Letiště :12 14:12–17:12) +1
Plzeň bus 21 every 15 (Bory :10 14:10–16:10, Litice :05 14:20–16:20) +1
Plzeň bus 25 / 24 2 an hour (25 Bory :07/:37 14:07–16:07) +1 each
Plzeň bus 26 and 42 26 every 60 :20 :40 / :04 :44 (14:20–16:04), 42 every 20 (14:30–15:50 / 14:34–16:14) +1 each
Plzeň bus 34 every 15 (14:03–15:40) +0–2 (2-min turns)
Plzeň bus 30P every 7–8 13:29–15:16 (shift-change peak; 61-min run) +5 or more
Plzeň bus 41K 4–5 an hour 14:05–15:37, irregular +3
Plzeň trolleybus 12 / 15 / 14 12 every 6 (14:21–15:27); 15 5 an hour; 14 2 an hour before 15:15 +2 / +1 / +2
Brno trolleybus 34 every 10 (Česká :08 15:18–16:18, Vychodilova :00 15:30–16:00; gaps at 15:08 and 15:20) 0 (same 4)

6. Verification

  • Rebuilt with python -X utf8 scripts/run.py --state : brno-trolleybus, brno-bus, plzen-trolleybus, plzen-bus, pardubice. Brno tram and Plzeň tram not rebuilt (no change, and no code they load changed; their pages are identical to this morning's).
  • htmldiff.py against this morning's pages: only the rows above and their notes, plus the Plzeň bus pak sentence (checkout list) and the method-text wording; against _snapshots/2026-10-25 the offset-fix rows in addition.
  • city_systemic.py: RATE, ORIGIN, RANGE, FOLD = 0 on all six GTFS sheets (rerun after the last rebuild). Approximate counts: Brno bus 4, Plzeň bus 14 (was 6), Plzeň trolleybus 11 — the Plzeň bus increase is labelling: the new rows are exact over a longer stretch of the peak (shown on the sheet as "partial, exact "), the old ones were exact only inside the 16:00 hour.
  • Pardubice (own check): RATE 0 on every non-daily line; ORIGIN/RANGE/FOLD 0. Hand-set rows no longer print the "merged minutes" line (their note says what is not modelled).
  • Not done: the README's Playwright look at 1400/390 px (only text cells changed).
  • Code changed (all pipeline-local, no shared common.py touched; brno-trolleybus only got a state override): brno-bus/scripts/03_timetable.py (override passing, stretch flag), brno-bus/scripts/05_render.py (label), plzen-trolleybus/scripts/03_timetable.py + 05_render.py (stretch flag and label; used by plzen-bus too), pardubice/scripts/03_network.py (fit_overrides) + 05_render.py (approx label, row note). READMEs updated.