[tex-live] Neo4j and TeX Live History

[Norbert Preining]

Dear all

I have now completed (but not finished) an import of each TeX Live
Database (texlive.tlpdb) since 2009-05-24 till today into a Graph
Database (Neo4j).

Represented entities are
- TLPDB
- TLPOBJ (so call TL Packages, that is what you install via tlmgr)
- TLPSRC (the source of those)
- Files (all files contained in a package)
- dependencies between all TLPOBJ (depend fields in tlpdb)
- contains: which TLPOBJ are contained in which TLPDB
- includes: which Files are contained in which TLPOBJ

Please see
https://github.com/TeX-Live/tl-neo4j
for a bit more details. There is also a read-only access to this
database at
http://texlive.info:7474/browser/
with user name and password both being "neo4j"

Just to give you an idea what can easily be done:

Find the packages with the highest number of updates over these 9+
years:

match (tlpdb:TLPDB) --> (p:TLPOBJ:Package)
with p.name as pkg,count(DISTINCT p)-1 as updates
where updates > 0
return pkg,updates order by -updates

This returns the following list, cut off at 100 updates:
pkg updates
bidi 185
dvips 172
tex4ht 152
context 134
siunitx 133
xepersian 125
pstricks 110
newtx 104
(here I restricted the search to TL Packages of type "Package", this
excludes "Context" packages as well as "TLCore" packages. The absolute
"winner" in terms of package updates is
texlive-docindex 1810
texlive.infra 945
...

All the best

Norbert

--
PREINING Norbert http://www.preining.info
Accelia Inc. + JAIST + TeX Live + Debian Developer
GPG: 0x860CDC13 fp: F7D8 A928 26E3 16A1 9FA0 ACF0 6CAC A448 860C DC13