Repository navigation
Releases: jeapostrophe/boku-ps1
Release list
Boku no Natsuyasumi (Japan), English v1
An English translation patch for the PlayStation original of Boku no Natsuyasumi
(SCPS-10088, 2000). Version 1, built from commit fe807a34998b0dd88e933a467a92842104acd600 (tag v1).
This is a patch, not the game. You need your own dump of the Japanese disc; nothing here
contains it, and nothing here is ever sold.
What is in English
The build wrote 2,983 lines of English into the game. The translation's 40 units (translation/status.tsv): 2 reviewed, 38 checked. The voices stay Japanese on purpose: this is a translation with subtitles, not a dub.
Download
boku-ps1-v1.zip is the download, under the names the instructions below use:
Boku no Natsuyasumi (Japan) [T-En boku-ps1 v1].xdelta— the patch (xdelta3).Boku no Natsuyasumi (Japan) [T-En boku-ps1 v1].cue— the cue sheet for the patched image.PATCH.jsonandREADME.txt— both sides' hashes, for a script and for a person.
What you need
A raw dump of the Japanese disc, one MODE2/2352 track, matching Redump (http://redump.org/disc/4890/). Hash it
before you patch: if these do not match, the patch is not for your file.
| File | Boku no Natsuyasumi - Summer Holiday 20th Century (Japan).bin |
| Size | 658,959,840 bytes |
| CRC32 | 39d1fe3f |
| MD5 | ee044864753c75ce0aea4ba777735fcd |
| SHA-1 | 5959bf7d9835d0a60aeb0143e2d0fc564bfea9fa |
How to apply it
-
Extract your CHD to a raw image (chdman
ships with MAME;brew install rom-toolson macOS):chdman extractcd -i "your dump.chd" -o "Boku no Natsuyasumi - Summer Holiday 20th Century (Japan).cue" -ob "Boku no Natsuyasumi - Summer Holiday 20th Century (Japan).bin"A BIN/CUE dump needs no extraction; use its
.bin. Never patch a.iso(2048-byte
sectors) or the.chditself. -
Hash the image and compare its SHA-1 with
5959bf7d9835d0a60aeb0143e2d0fc564bfea9fa:# macOS shasum -a 1 "Boku no Natsuyasumi - Summer Holiday 20th Century (Japan).bin" # Linux sha1sum "Boku no Natsuyasumi - Summer Holiday 20th Century (Japan).bin" # Windows certutil -hashfile "Boku no Natsuyasumi - Summer Holiday 20th Century (Japan).bin" SHA1 -
Patch it with xdelta3 (
brew install xdelta
on macOS; MultiPatch is a macOS app for the same
job):xdelta3 -d -s "Boku no Natsuyasumi - Summer Holiday 20th Century (Japan).bin" "Boku no Natsuyasumi (Japan) [T-En boku-ps1 v1].xdelta" "Boku no Natsuyasumi (Japan) [T-En boku-ps1 v1].img"The browser patcher RomPatcher.js cannot apply this patch.
-
Play: put
Boku no Natsuyasumi (Japan) [T-En boku-ps1 v1].cuebesideBoku no Natsuyasumi (Japan) [T-En boku-ps1 v1].imgand open the.cuein your emulator. You
can recompress withchdman createcd -i "Boku no Natsuyasumi (Japan) [T-En boku-ps1 v1].cue" -o "patched.chd".
Optional: the PPF, for DuckStation
While the release page lists it, boku-ps1-v1-ppf.zip is a separate download holding the same patch as a
PPF, with its own PATCH.json, README.txt and the cue sheet. DuckStation applies it
without patching anything: give Boku no Natsuyasumi (Japan) [T-En boku-ps1 v1].ppf your CHD's name with .ppf in place of .chd
(Boku.chd → Boku.ppf), put it beside the CHD, and tick Settings → CD-ROM → Apply Image
Patches, which is off by default. DuckStation checks nothing, and the hashes above are of
the extracted image, not the CHD — extract it once (steps 1 and 2) to check your dump.
What you should get
A patched image whose SHA-1 is 3ab94dcb5338bd17a8cf848d931efc0476fc9aa4 is v1, whatever the file is called.
The README on the project's GitHub page lists every version's, under "Which version do I have?".
| File | Boku no Natsuyasumi (Japan) [T-En boku-ps1 v1].img |
| Size | 658,959,840 bytes |
| CRC32 | 01040247 |
| MD5 | 1990984d00fe36b8d05ce2640a65d47a |
| SHA-1 | 3ab94dcb5338bd17a8cf848d931efc0476fc9aa4 |
README.txt in the download says all of this again, including what the patch does and does
not check for you.
Where it is played
The project is developed and played on Beetle PSX (the mednafen core, as in RetroArch) and
DuckStation. It has never been tested on a real PlayStation.
How the translation was made
The English script is written by AI agents — Claude, the strongest model available, under
Jay's direction — not a sentence-at-a-time machine translation. The game's event scripts were
decoded so that a translator sees a whole scene as the game plays it — which line follows
which, where the choices branch, who is speaking — after reading a story bible, a style guide
and a glossary written for this project. A second agent reviews every line against the
Japanese, Jay reads the script and rules on the questions it raises, and then it gets played.
The reverse engineering and the assembly patches that make room for English were done with
Claude too. The project is made for Jay and his friends and published for anyone who wants it;
everything that produced the script is in the repository to inspect, and pull requests that
improve it are welcome.
Credits
None of these target the PS1 original's text, and all of them made this project's first day
shorter. What each one knows is written up in
research/related-projects.md.
- pleonex/Boku-no-Natsuyasumi — PSP port
extraction tools and a format wiki; its glyph table is what first decoded the PS1 script. - HilltopWorks/BokuNoNatsuyasumi2 — the
PS2 sequel's complete English patch toolchain. - GriffithVIII/Boku-no-Natsuyasumi-ESP —
the PSP port's Spanish patch (TraduSquare). - psyouloveme/boku1-reversing — PS1 original:
Ghidra scripts, a jPSXdec index of the disc, and a RAM map from speedrunning. - KendritPy/Boku_ESP_JP and
snake7594/boku-natsu-portable-kr-patch —
PSP port: a precise dialogue-format spec, and a 2,020-code character table.
Licence
The tools and patches' source are MIT; the English script and its context notes are
CC BY-SA 4.0 — both in the source archive attached to this release. These patches were
built from a clone of the repository at this release's tag, with ./make.sh import on a
dump and then ./make.sh release. Boku no Natsuyasumi is © Sony
Interactive Entertainment; this project distributes none of it and is not affiliated with
Sony or Millennium Kitchen. Paths in code above are files in the source archive.