ACIP Uncatalogued Archive — Survey

What is in the ACIP Dropbox, before we catalogue it · 21 August 2026

A metadata-only inventory of the ACIP Dropbox (no files downloaded) — the fruit of ~30 years of Geshe Michael’s project to find, rescue and digitise Tibetan texts. Almost the entire archive is digitised text in ACIP’s native format (whole-volume .ALL files, per-work .ACT files, and numbered .001–.NNN page segments) — only 36 files are scans.

210,904
files (8.3 GB, 12,616 folders)
36
scans (PDF/image) — the rest is text
1%
exact-byte duplicate copies
~24
author-collection works NEW to our catalogue

The five top-level libraries

The Dropbox has five top folders. The catalogue work below concentrates on the two that hold directly-catalogable collection material; the other three are Gene Smith’s intros, and the large canon/reference/raw-data layers.

Library (top folder)FilesSizeWhat it isRole
JASON DROPBOX AS OF 8 1253,2474.40 GBAuthor collection volumes + a raw-data foldertier 1 (the unsplit volumes)
NEWDATA603157,0153.64 GBKangyur in ACIP format + a large Reference library + old data dumpscanon + reference, not new Sungbum
Input Orders from 2015 on4850.23 GBWhole-collection .ALL volumes by input batchtier 1
acip unsplit volumes as of 9 111320.04 GBAuthor catalogues + page segmentstier 1 (partial)
GENESMITHINTROS250.00 GBGene Smith’s encyclopedic intros (.ALL)catalogable, small

The three tiers

1a · Collection volumes to split — where vol3/vol4 live

The Input Orders from 2015 on folder holds 360 whole-collection .ALL volumes, filed by monthly input batch. This is exactly where the vol3/vol4 work came from — e.g. 23 OCTOBER 2017/VEN RENDAWA VOL 3.ALL and VOL 4.ALL — alongside many more collections (Tsarchen, Milarepa Sungbum, Nepal Library…), each one ready to split and catalogue the way we did the Rendawa volumes.

Input batch.ALL volumes
43 AUGUST 201925
33 AUGUST 201819
09 JUNE 201618
32 JULY 201818
10 JULY 201617
07 APRIL 201615
34 SEPTEMBER 201815
31 JUNE 201814
05 SEPT 201513
29 APRIL 201812
58 JANUARY 202111
30 MAY 201811
35 OCTOBER 20189
38 JANUARY 20199
52 JULY 20209
67 JANUARY 20229
08 MAY 20168
55 OCTOBER 20208
62 AUGUST 20218
06 MARCH 20167
59 FEBRUARY 20217
66 DECEMBER 20217
21 AUGUST 20176
39 FEBRUARY 20196
23 OCTOBER 20175
72 JUNE 225
26 JANUARY 20185
19 JUNE 20174
57 DECEMBER 20204
69 MARCH 20224
81 APRIL 234
45 OCTOBER 20194
16 MARCH 20173
22 SEPTEMBER 20173
24 NOVEMBER 20173
28 MARCH 20183
36 NOVEMBER 20183
14 JANUARY 20173
37 DECEMBER 20182
48 FEBRUARY 20202
68 FEBRUARY 20222
78 JANUARY 232
80 MARCH 232
13 DECEMBER 20162
04 AUG 20151
18 MAY 20171
20 JULY 20171
27 FEBRUARY 20181
54 SEPTEMBER 20201
56 NOVEMBER 20201
60 MARCH 20211
61 JULY 20211
65 NOVEMBER 20211
79 FEBRUARY 231
Misc1
44 SEPTEMBER 20191
64 October 20211
71 MAY 20221

1b · Author collection works

The ACIP unsplit volumes folders: 2,587 works (.ACT/.ALL) organised under 74 author-folders, each with BDRC W-numbers and a cat.txt. Cross-checked by ACIP S-id against our Sungbum catalogue: 1053 already held, ~24 genuinely new.

2 · Kangyur in ACIP format

NEWDATA603/ORANEW/Kangyur — 28,125 files: the canonical Kangyur in native ACIP format (volumes + page segments). We already hold most of this as our Lhasa Kangyur.

3 · A large reference library

NEWDATA603/ORANEW/Reference — 124,686 files: a very large additional corpus (catalogues and texts). Worth a separate survey before any cataloguing.

By top-level folder

FolderFilesSize
NEWDATA603157,0153.64 GB
JASON DROPBOX AS OF 8 1253,2474.40 GB
Input Orders from 2015 on4850.23 GB
acip unsplit volumes as of 9 111320.04 GB
GENESMITHINTROS250.00 GB

Author folders in the collection volumes (74)

Author folderWorks
GM-STRUCTURE1,019
dharma-bhadra474
blo-bzang-rta-mgrin295
a-khu-ching-shes-rab-rgya-mtsho120
blo-bzang-nor-bu-shes-rab72
blo-bzang-bsam-grub69
dalai05-ngag-dbang-blo-bzang-rgya-mtsho43
blo-bzang-rab-gsal34
blo-bzang-'jam-dbyangs-smon-lam28
che-shos28
'ju-mi-pham27
((non-acip-inputs))+24
thu'u-bkwan-3-blo-bzang-chos-kyi-nyi-ma22
yongs-'dzin-ye-shes-rgyal-mtshan19
dge-'dun-chos-skyong17
grags-pa-bshad-sgrub16
((multiple-author-collections))14
khri-byang14
dalai02-dge-'dun-rgya-mtsho13
dalai07-skal-bzang-rgya-mtsho13
kirti08-blo-bzang-'phrin-las13
sle-lung-blo-bzang-'phrin-las12
mkhas-grub-sangs-rgyas-ye-shes11
dalai01-dge-'dun-grub11
brag-dkar-blo-bzang-dpal-ldan10
rgyal-tshab9
dkon-mchog-'jigs-med-dbang-po9
mnyam-nyid-rdo-rje9
panchen01-blo-bzang-chos-rgyan9
lcang-lung-pandita8
tsong-kha-pa6
cha-har-dge-bshes6
mkhas-grub5
'jig-rten-mgon-po5
chu-bzang-ye-shes-rgya-mtsho5
dbyangs-can-grub-pa'i-rdo-rje5
dkon-mchog-rgyal-mtshan5
lcang-skya-3-rol-pa'i-rdo-rje5
o-rgyan-'jigs-med-dbang-po5
pha-bong-kha5
Recommendation. Start with tier 1 — have the Captain copy only tier 1 (the small, high-value slice) into a new, separate uncataloged library area — that is the 360 .ALL collection volumes in Input Orders plus the 2,587 author works. Then split the .ALL volumes and process the ~24 new works exactly as we did for vol3/vol4. Tiers 2–3 are large; survey them separately rather than copying 8 GB up front. Duplicates are only 1% by exact bytes, so little is wasted — but content-hashes let us skip any copy on ingest.
Metadata-only inventory (no downloads) generated 21 August 2026; 210,904 files fingerprinted by Dropbox content-hash. Self-contained — opens offline, day/night toggle.