tianjin-europe-2026

Date	From → To	Mode	Time	Price (USD equiv)	Status
Wed Jul 8	Tianjin TSN (or Beijing PEK) → Paris CDG	Flight	~14h, 1 stop	~$464	booked
Thu Jul 9 — Mon Jul 13 (+Tue Jul 14 morning)	Paris (5 nights, mandatory stop)	—	—	—	booked · paris-hotels.html
Tue Jul 14	Paris BVA → Budapest BUD (Ryanair FR4230)	Flight (low-cost)	2h 10m direct	low-cost (on booking)	booked
Tue Jul 14 — Fri Jul 24 (10 nights, base)	Budapest (+ optional day-trips to Vienna/Bratislava)	Hotel	—	—	hotel TBD
Fri Jul 24 evening	Budapest Nyugati → Warszawa Centralna (default; or Vienna/Bratislava)	Overnight sleeper (EN Metropol)	~12h 30m	~$55–145	open · sleeper options
Sat Jul 25 morning	Arrive Warszawa Centralna ~08:00	—	—	—	target

Option	Routing	Duration	Price (USD)	Trade-offs
1. PEK direct (recommended for simplicity)	Air China CA875 or Air France PEK→CDG, nonstop	~12h45m	$700–900 (est, unverified)	Simplest. 1 PNR. Add TSN→PEK domestic (~1h, $60–80) if starting from Tianjin — separate booking, short connection. Trip.com sampled Air China nonstop from $726 one-way (Jul 21) — Jul 8 likely similar or slightly higher.
2. TSN 1-stop via PEK (single PNR)	Air China TSN→PEK→CDG	~14h total	$760–980 (est, unverified)	Single PNR with checked bags. Longer than PEK direct because of domestic leg. Real cost = nonstop PEK→CDG + domestic TSN→PEK sector.
3. TSN 1-stop via Gulf	Qatar / Emirates / Etihad via DOH / DXB / AUH	~16–20h	$450–700 (est, unverified)	Likely cheapest. Long travel day with 3–5h layover at a Gulf hub. Skyscanner's "from $311 / $365" route-page lows are NOT verified for Jul 8 — could be off-peak or already sold out fare classes.

Log (newest first)

2026-07-08 16:00 UTC · budapest-hotels.html sub-page created (solo + own room + no kitchen)

ishmael: "I am solo, I want my own room, don't want a kitchen. make a budapest-hotels.html". New sub-page budapest-hotels.html mirroring the structure of paris-hotels.html: TL;DR + 2 option tables (hostel private rooms, mid-range hotels) + district guide + recommendation + booking workflow + data sources + log. The "no kitchen" rule kicked out the entire apartment tier (Elite Studios, Agape Guesthouse, all Booking/Airbnb apartments in VII) — leaving hostels-with-private-rooms + mid-range hotels in V/VI/VII/VIII. Top 3 picks: Maverick Lodges private room (~$45/nt, social/price), Bohem Art Hotel (~$60/nt, quiet/hotel-quality), Hotel Palazzo Zichy (~$65/nt, value-for-quality). All free-cancellation on Booking. Curated from 2 local guides (hotelierschoice.com + budapestbylocals.com) — live Booking prices for Jul 14-24 are not on the page (visible-chrome scrape would need to bypass the Booking consent + date-picker UI), the Booking links in the tables return live prices for the specific dates. Files section + Next-steps #4 now point at the new page.

2026-07-08 14:30 UTC · pivot to Budapest — Ryanair FR4230 BVA→BUD booked, V/B/Br 12nt block collapsed to 10nt Budapest

ishmael: "ok so I booked ryan air FR4230 2026-07-14 leaving 19:25 / which means I'll also be staying a night longer in paris (at the ibis)". Big change. FR4230 is the Ryanair Paris-Beauvais → Budapest route (BVA→BUD), Tue Jul 14 19:25 → 21:35 (2h 10m). This is not a Paris → Vienna leg — it's a Paris → Budapest leg. The whole "12 flexible nights in Vienna/Budapest/Bratislava" block has collapsed: Budapest is now the single base for 10 nights (Jul 14-24), with Vienna and Bratislava downgraded from "competing base candidates" to "optional day-trips or 1-2 night side-trips". Paris extended from 4 to 5 nights (ibis, Jul 9-14). What this unlocks: (a) the 12nt V/B/Br distribution question is gone — no more "which city gets how many nights" — Budapest just gets all 10. (b) the sleeper option decision is much simpler — option B (Budapest Nyugati 19:30, EN Metropol) is the obvious default. Vienna/Bratislava options remain on the table only if a last-minute side-trip becomes appealing. (c) the Paris→Vienna flight is no longer needed (the previous "open: book CDG→VIE" todo is dead). Edits: subheader, meta description, "Updated" timestamp, full TL;DR (pivot callout at top + rewritten), plan table (Paris 4nt→5nt row, CDG→VIE row replaced with BVA→BUD row, V/B/Br 12nt block replaced with Budapest 10nt base, sleeper row's status updated to "option B default"), Next-steps (replaced CDG→VIE todo with BVA→BUD + Budapest-hotel todos, reshuffled sleeper-window todo to mention MÁV for option B), Files section (updated paris-hotels.html description to note 5th night), home page card (updated blurb), and this log entry. What did NOT change: the Hainan via-XIY outbound is still locked ($464), Paris hotel is still booked (now 5 nights at the same ibis), the warsaw-sleeper-options.html sub-page is still accurate as the data reference for all 3 boarding options. Open question for ishmael: is the Budapest base for all 10 nights the final plan, or do you want a Vienna/Bratislava side-trip on top?

2026-07-05 05:55 UTC · paris → vienna flight price was unverified, fixed

ishmael: "where did you find a $100 paris -> vienna flight, I can't find any that cheap". Right to push back. The ~$90–130 / ~€80–120 number I'd written on the project index, itinerary.html, and flight-research.html was a guess based on a vague sense of EU short-haul pricing, not a verified live price. Replaced it everywhere with the actual Google Flights number for Mon Jul 13 (scraped 2026-07-05 via headed chrome on :1): cheapest €146 OW / ~$170, mid-tier €171–182. That's about 30–50% higher than my old guess. Fixed in 3 files: project index (plan table row 3 + next-steps list), itinerary.html (Leg 3 blurb), flight-research.html (the long-haul section's "not yet verified" footnote). Lesson: don't put dollar figures on open-decision legs of the plan. Either verify live and cite the source, or mark it as an estimate. Fixed in the 3 places I'd put a number on a not-yet-booked leg. Note 2026-07-08: this todo is now obsolete — the Paris→Vienna leg was replaced with a Paris→Budapest Ryanair flight (FR4230, see the 2026-07-08 14:30 entry above).

2026-07-05 05:40 UTC · round-2 cleanup — slim plan table, drop Quick facts from itinerary

ishmael: "tables should be short and sweet" + "itinerary.html does not need a quick facts". Fixed:

Plan table cells slimmed. The Status column was the problem — row 4 (V/B/Br open block) had an 825-character cell with the entire 6-option reference list stuffed inside one cell, which made the table hard to read. Replaced it with a one-liner <span class="tbd">open</span>. Same treatment on rows 1–3 and 5: status cells are now booked / booked · paris-hotels.html / open / open · sleeper options. Hotel name, address, and the full sleeper-window discussion moved out of the table.
6 reference shapes moved below the table as a follow-up list, not inside the V/B/Br row. That separates "here's the locked plan" (the table) from "here are the open decisions and their plausible shapes" (the list below). No more night counts in the shapes themselves — "Vienna → Budapest → Bratislava, ends Bratislava (board option C)" instead of "Vienna 4 + Budapest 7 + Bratislava 1".
Quick facts section removed from itinerary.html — the Total days / Countries visited / Paris stop / Locked arrival / Open bullets were restating what the legs already say. The page is now shorter and the table is the single source of truth for the chain.

Same fix spirit also applied to rows 1, 2, 3, 5 of the plan table — every status cell is now ≤30 chars. After-fix verification: 0 cells in the plan table exceed 60 chars. Plan table is 6 rows, each row scannable in a single glance.

2026-07-05 05:30 UTC · itineary.html cleanup — day count, prices, status section, V/B/Br night counts

ishmael: "8-25 is def not 25 days" + 4 other cleanup points all on itinerary.html. Fixed:

Day count corrected from 25 to 18 (Jul 8 → Jul 25 inclusive is 18 days, not 25). H1, meta, "Days" badge, Quick facts — all updated on itinerary.html. Same fix on index.html H1, meta, TL;DR, and the home page card.
Prices stripped from itinerary.html — removed the totals card (3 cards: Flights $555-650 / Trains $70-200 / Total $625-850), the per-leg Price column on Legs 1, 3, and 5, and all €/$ references in the leg notes. Replaced the totals card with a one-liner pointing to project status + sleeper options pages for cost tracking. Sleeper sub-page (warsaw-sleeper-options.html) still has prices — it's a cost research page, that's its job.
Outbound flight collapsed to a one-liner on itinerary.html (no more $464 / Air China / Air France alternatives — the flight is already booked; further options live on flight-research.html).
Specific V/B/Br night counts removed from the 6 reference shapes (α-ζ) — now "Vienna → Budapest → Bratislava, ends Bratislava (board option C)" instead of "Vienna 4 + Budapest 7 + Bratislava 1". The shapes are sketches, not commitments.
Status h2 section removed from itinerary.html entirely. (The project status pill in the header still says wip, that's the right place for it.)

Same 18-day fix deployed on the project index (H1, meta, TL;DR) and home page card. Deploy verified: headless chrome renders, grep confirms 0 occurrences of "25 days" / "$464" / "€80" / "Air China" / "Vienna 4 + Budapest 7" / "<h2>Status</h2>" on the live itinerary.html.

2026-07-05 04:57 UTC · overhaul — warsaw target pulled in to sat jul 25, V/B/Br now open

ishmael: "ok update the webpage(s). do an overhaul - there's a lot of information here that seems to be out of date." Current state of the trip, as ishmael restated it: Wed Jul 8 flight with layover in XIY · Thu Jul 9 land in Paris · booked hotel, stay till Mon Jul 13 · Paris → Vienna (transit still unbooked) · Vienna/Budapest/Bratislava undecided · overnight sleeper to Warsaw arriving Sat Jul 25. This is a major pivot from the previous "Sun Aug 2 target + Bratislava 6nt + NJ 471 Thu Jul 30 night" plan — the whole second half (Bratislava 6nt, NJ 471 Thu Jul 30, Warsaw 2nt buffer) is gone. The new invariants: arrive Paris Thu Jul 9 (hotel locked), board sleeper Fri Jul 24 evening, arrive Warsaw Sat Jul 25 morning. Everything in between (12 nights, Mon Jul 13 → Fri Jul 24) is a flexible 3-city block: which cities, in what order, for how long. Edits to index.html: subheader, meta description, "Updated" timestamp, full TL;DR (replaced), Files section (added the new warsaw-sleeper-options.html sub-page), the entire plan table (replaced 12 rows with 6 rows: outbound, Paris, Paris→Vienna, the new open 12nt V/B/Br block, sleeper, Warsaw arrival), Next-steps (replaced), and this log entry. New sub-page: warsaw-sleeper-options.html sub-page), the entire plan table (replaced 12 rows with 6 rows: outbound, Paris, Paris→Vienna, the new open 12nt V/B/Br block, sleeper, Warsaw arrival), Next-steps (replaced), and this log entry. New sub-page: warsaw-sleeper-options.html — 3 boarding options (Vienna / Budapest / Bratislava) with full seat61-derived times, fares, and booking-window status. Edits to itinerary.html: subheader, meta, header status, totals card (trains: BUD↔VIE and BUD↔Bratislava only — no more 4-leg train chain), the entire 12-leg itinerary table (replaced with 5 legs: outbound, Paris, Paris→Vienna, the open 12nt V/B/Br block, sleeper, end), Quick facts list, and Status block. Edits to srv/index.html: tianjin-europe-2026 card blurb updated to reflect Sat Jul 25 target and open V/B/Br distribution. 6 reference V/B/Br splits (α through ζ) in the plan table — none of them are decided, they're just a menu of plausible shapes the 12nt block could take. Status pill still "wip" — the V/B/Br split, the sleeper boarding choice, and the Paris→Vienna transit are still open. What did NOT change: the Hainan via-XIY outbound is still locked ($464), Paris hotel is still booked, the outbound flight-research.html and paris-hotels.html pages are still accurate. Note 2026-07-08: this whole overhaul is now superseded by the Budapest pivot — see the 2026-07-08 14:30 entry at the top of the log.

2026-07-04 07:50 UTC · paris hotel booked, plan shifted +1 day, vienna compressed 5→4nt

ishmael booked ibis Paris la Villette Cité des Sciences 19th (31-35 quai de l'Oise, 75019 Paris) for Jul 9-13, 4 nights. Two realignments followed:

Paris block shifts to Jul 9-13 (was Jul 8-12) because the Hainan via-XIY layover makes Wed Jul 8 a transit day at XIY T3, not a Paris day. ishmael confirmed "jul 8th will be at the xiy airport" and "update it so that I leave for vienna on the 13th".
Vienna compressed 5 → 4 nights (Mon Jul 13 → Fri Jul 17) to keep the Sun Aug 2 Warsaw target — Budapest, Bratislava, NJ 471 Thu Jul 30 night, and Warsaw 2nt all unchanged.

ibis La Villette caveat: the hotel wasn't in the verified Paris research (the research was for central Paris private rooms; Booking's chain-brand URL filter for ibis doesn't work, so the ibis prices cited on paris-hotels.html are direct-from-Accor estimates, not Booking-verified). ishmael made the booking himself and didn't ask for a comparison. Hotel is 19th arr, ~5-6 km from central Paris, near Porte de la Villette / Cité des Sciences (metro line 7) — fine for the role, just not the central-Paris profile the research was scoped to. Edited: subheader, meta timestamp, TL;DR, plan table (Paris / Paris→Vienna / Vienna rows), long-haul callout text, "Book CDG→VIE" todo, "Book hotels" todo, and added the booking details to the Files section. paris-hotels.html also updated with a "booked" callout and a corresponding log entry. Status pill still "wip" — long-haul outbound, CDG→VIE, NJ 471 inventory, and 4 city hotels (Vienna, Budapest, Bratislava, Warsaw) still TBD. Note 2026-07-08: the 4-night Paris block was extended to 5 nights and the Vienna leg was replaced by a Budapest leg in the 2026-07-08 14:30 pivot entry.

2026-07-02 04:30 UTC · reverted 03:55 bratislava-drop, restored 09:45 plan (Paris 4nt + Bratislava 6nt + Budapest 7nt + Warsaw Sun Aug 2)

ishmael: "fix it". The 03:55 deploy misread the chain: the assistant heard "warsaw target can never move" and assumed the target was the original Sat Jul 25, so it dropped Bratislava and cut Budapest to make Sat Jul 25 fit. But the canonical 09:45 plan's target was already Sun Aug 2 (8 days later than Sat Jul 25, the result of the 4-night Paris extension). Reverted to the 09:45 plan: Paris 4nt Jul 8–12, Vienna 5nt Jul 12–17, Budapest 7nt Jul 17–24, Bratislava 6nt Jul 24–30 restored, NJ 471 Thu Jul 30 night, Warsaw 2nt Jul 31–Aug 2, target Sun Aug 2. Edited: title, subheader, meta description, "Updated" timestamp, TL;DR, the 5 missing Bratislava-leg rows in the plan table (12 rows again), Next-steps "Verify NJ 471" and "Book hotels" todos, and the paris-hotels.html description line in the Files section. The 03:55 entry below is left intact per AGENTS.md chronological-log rule — it describes what was on the page between 03:55 and 04:30, which is a true historical state even if it was wrong.

2026-07-02 03:55 UTC · bratislava dropped, warsaw target held at Sat Jul 25 (REVERTED 04:30 — see entry above)

ishmael: "the warsaw target can never move. I definitely don't care about bratislava as much, and ideally I would arrive in warsaw right on time." Cut the chain to make Sat Jul 25 feasible while keeping Paris 4 nights (Jul 8–12): dropped Bratislava entirely (0 nights, was 6) and cut Budapest 7 → 5 nights (Jul 17–22, was Jul 17–24). New chain: Paris 4 + Vienna 5 + Budapest 5 + Warsaw 2 = 16 nights + 2 transit (Paris→Vienna, Bp→Vienna) + 1 NJ = 19 days on the ground, fits inside Wed Jul 8 → Sat Jul 25. NJ 471 now Wed Jul 22 night (was Thu Jul 30 night in the previous revision). Edited: subheader, meta description, TL;DR, full plan table (12 rows, Bratislava leg rows removed), long-haul section, "Verify NJ 471" todo, "Book hotels" todo, and "Updated" timestamp. Old log entries that still reference Bratislava left intact per AGENTS.md chronological-log rule. 2026-07-02 04:30 UTC · REVERTED: this revision was wrong (see entry above). Note 2026-07-08: the whole Vienna/Budapest/Bratislava block was later collapsed to a single Budapest base in the 2026-07-08 14:30 pivot entry.

2026-07-01 09:45 UTC · trip dates revised + index page brought current

ishmael said "the index page is still out of date" and "I want to leave on the 12th not 11th". Updated the project plan to Paris 4 nights Jul 8–12 (was 3 nights Jul 1–4) and shifted every downstream leg by 1 week + 1 day: Vienna 12–17, Budapest 17–24, Bratislava 24–30, NJ 471 Thu 30, Warsaw 31–Aug 2, target Sun Aug 2 (was Sat Jul 25). Outbound flight target is now Wed Jul 8 (the $464 Hainan via XIY already-verified option from the 2026-06-30 comparison window). Edited: subheader, meta description, TL;DR, full plan table (12 rows), and the "Book hotels" todo. Did not retro-edit the older log entries that still reference Jul 1 — those describe what was true when written, per AGENTS.md chronological-log rule.

2026-07-01 09:30 UTC · chain options added to paris-hotels.html

ishmael asked for the link to the research page (he'd asked twice across the session), and when I sent it, he pointed out it didn't include the 3-star chain data from the followup conversation. Added a new "Option C: 3-star chain hotels" section to paris-hotels.html with 16 Astotel + 4 Campanile + 8 HappyCulture properties (live property lists scraped from the chains' own websites; live prices not included because Astotel and HappyCulture have no real-time booking widget and Campanile's rate widget needs session cookies). Also added a 7-row "Other 3-star chains" table (B&B, Hipotel, Citadines, ibis, Kyriad, Best Western, Mercure). Link-target audit passed (74 links, 0 violations).

2026-07-01 04:11 UTC · paris accommodation research

ishmael asked for Paris accommodation Jul 8-10 or 8-11, solo, $100/night, central. Live-scraped Booking (2-night + 3-night), Hostelworld, Airbnb on :1. Key finding: $100/night in central Paris = hostel dorm bed, full stop. Cheapest private rooms start $130-180/night. Top picks: Art Beaubourg dorm $87/nt (3rd arr, 0.9 km); The People Marais dorm $69/nt (4th arr, 1.1 km, score 8.1); Au Fleuri private $135/nt (20th arr, 3.8 km, score 5.3). New sub-page: paris-hotels.html.

2026-06-30 11:55 UTC · outbound comparison window

Per ishmael's 2026-06-30 request ("I want to leave like the 7/8/9th or so"), re-scraped Beijing → Paris / Vienna outbound OW for Tue Jul 7 / Wed Jul 8 / Thu Jul 9. 64 rows total, all OW verified in headed chrome on :1. Cheapest per route: PEK→CDG Jul 7 = $464 Hainan via XIY (25h 55m layover, bad timing — same route as Jul 1) / Jul 8 = $464 Hainan via XIY (7h 30m overnight, same as Jul 1, the cleanest deal of the three days) / Jul 9 = $479 Hainan via SZX (4h 45m). PEK→VIE all $333 Hainan via TFU (long overnight; sometimes requires CTU↔TFU airport change mid-layover, ⚠ avoid). The new section in flight-research.html replaces the previous "Return Jul 7/8/9" table (which had the wrong direction). Raw CSV: 2026-06-30-beijing-paris-vienna-OW-jul-7-9.csv. Trip plan is still Jul 1 outbound, this is a comparison window for if you wanted to delay by a week.

2026-06-29 10:35 UTC · verification

Long-haul outbound flight prices verified via Google Travel Explore + Google Flights (headed chrome on :1). Trip is one-way; only outbound Jul 1 was scraped. Earlier "captcha-gated" caveat was wrong — the consent screen is the gate, not an IP block. Earlier "round-trip" framing was also wrong — the trip plan ends in Warsaw, no return to China. The "$964 RT Hainan" was the same Hainan flight I now have as $464 OW. Cheapest outbound OW: $464 Hainan PEK→CDG via XIY (7h 30m overnight, in-airport), $333 Hainan PEK→VIE via TFU (⚠ second $333 Hainan option requires CTU→TFU airport change mid-layover, avoid). PP correction at XIY: my earlier "limited PP" was wrong — PriorityPass lists 6+ lounges across T2/T3/T5. Hainan at XIY uses T3, so V38 / Simple Business / China Southern VIP are easy. Full table + PP lounge ratings in the new flight-research.html sub-page; raw CSVs in /projects/google-travel-explore/. 2026-06-30 03:32 UTC · cleanup: removed the return-leg table from flight-research.html (this trip has no return flight). Deleted the return-leg CSV and the RT-OW comparison CSV from the project folder and mni.ac. Repositioned the page as outbound-only. The "$1,326 Paris RT" / "$801 Vienna RT" framing that was on the page was misleading for a one-way trip — the real outbound cost is just the single $464 or $333, not the sum.

2026-06-28 13:05 UTC · correction

Long-haul pricing was overstated. The "$310 via Gulf" claim was a Skyscanner cached aggregate, not a verified Jul 1 fare — user correctly flagged it. Real verified data: Trip.com's Jul 21 calendar shows Air China nonstop PEK→CDG from $726 one-way. Adjusted long-haul estimate to $450–900 depending on routing (nonstop $700–900, connecting $450–700). Updated the totals card on itinerary.html ($800–1,030 total transport) and the long-haul options table on index.html. Earlier "$610–810 / saves ~$110" claim retracted. TL;DR + status table updated.

2026-06-28 10:09 UTC · user message

User revised: drop Shenzhen + Hong Kong from the plan. Direct Tianjin/Beijing to Europe, same other requirements: mandatory 2–3 days Paris, then Vienna / Budapest / Bratislava, arriving Warsaw on Jul 25. Restructured the chain — Paris becomes the first stop (3 nights, arrives Wed Jul 1 evening); Vienna 5 nights; Budapest 7 nights; Bratislava 6 nights (no longer optional); Warsaw 2 nights buffer. Long-haul simplified to TSN/PEK → CDG with three pricing options. Original "saves ~$110" cost claim retracted 2026-06-28 13:05 — see correction entry above. Status page + itinerary page updated.

2026-06-25 08:56 UTC · user message

User revised: switch European order to Vienna → Budapest → [Bratislava] → Warsaw. Rationale: putting Bratislava between Budapest and Warsaw means dropping it cleanly reverts to the simple Budapest → Warsaw sleeper route without restructuring anything else. (Was previously Vienna → Bratislava → Budapest → Warsaw.) Superseded 2026-06-28: Bratislava is no longer optional; it's now a 6-night base between Budapest and Warsaw.

2026-06-25 08:40 UTC · user message

User corrected three things: (1) TSN → SZX is ~$110, not $410 (my Trip.com calendar read was wrong); (2) all EUR/HKD prices should show a USD approximation alongside; (3) drop Berlin from the European leg — lengthen Vienna to 4 nights and Budapest to 6 nights, and insert 2 nights in Bratislava between Vienna and Budapest. Last leg becomes Budapest → Warsaw overnight sleeper (replaces Berlin → Warsaw). Superseded 2026-06-28: Shenzhen + HK dropped; chain reshaped.

2026-06-25 08:11 UTC · project created

Created the project page + shareable itinerary page. Tentative plan locked: TSN → SZX (Jul 1) → HKG (1 night Jul 7) → CDG via ZRH on SWISS (Jul 8) → Paris (3 days) → VIE fly (Jul 12) → Vienna → Budapest → Berlin → Warsaw (Jul 25). European leg transport modes and bookings still pending. Superseded 2026-06-28: Asia legs dropped, Berlin removed.

2026-06-25 08:06 UTC · user message

User locked the tentative plan with the following changes from the FRA entry I'd proposed: drop FRA, switch to SWISS HKG → ZRH → CDG (LX139 + LX632); confirm Nightjet Paris→Vienna is no longer running and fly CDG→VIE on Jul 12; leave European leg routing open for now.

2026-06-25 07:16 UTC · verification

Verified live on ÖBB shop: Paris → Vienna daytime options €130–235 (no NJ 469 Nightjet shown — service may be suspended); Berlin → Warszawa Centralna overnight sleeper €42–62 with 4 verified departures all arriving Sat Jul 25 morning (target met). Berlin leg since dropped; Budapest → Warsaw sleeper needs re-verification (was Berlin → Warsaw originally).

2026-06-25 06:54 UTC · user pushback

User correctly flagged that my SZX→CDG "$350" estimate was wrong — pulled live Trip.com calendar lows instead. Real Jul 8 prices: SZX→FRA $300, SZX→CDG $421, SZX→VIE $445, SZX→BER $562. Lesson recorded: don't cite "from $X" route-page prices; use calendar lows.

2026-06-25 06:24 UTC · trip request

User requested multileg transportation plan: Tianjin → Shenzhen (+ optional HK) → Europe (mandatory Paris 2-3 day stop) → Warsaw on Jul 25 (literal arrival, not earlier). Preferred Vienna + Budapest; open to other cheap/interesting alternatives.

tianjin → europe, 18 days wip

TL;DR

Files

Current plan (full chain)

Long-haul options (TSN/PEK → CDG)

Next steps

Log (newest first)