update notes and explore

2 files changed, 26 insertions, 1 deletions

diff --git a/TODO.md b/TODO.md
index 29b1fe0..befbd48 100644
--- a/TODO.md
+++ b/TODO.md
@@ -1,4 +1,5 @@
 
+
 priorities:
 - coverage stats, particularly for longtail
 - `is_active` coverage
@@ -10,9 +11,22 @@ priorities:
 
 ## Sources
 
-- unpaywall journal-level classification
+- preservation coverage
+    x  hathitrust (huge!)
+        https://www.hathitrust.org/hathifiles_description
+    x  PKP PLN (ONIX)
+        https://pkp.sfu.ca/pkp-pn/
+        http://pkp.sfu.ca/files/pkppn/onix.csv
+    => Scholars Portal (canada)
+        received ONIX XML, hoping for KBART format
+    => Cariniana
+    => National Digital Preservation Program, China
+    => Library of Congress
+- additional hathitrust (many more ISSNs/journals)
+- unpaywall journal-level classification (OA color)
     => ask for journal-level dump or do munging
 - jurn matches
+    => somebody on github did an openrefine match
 - public scopus list (?)
 - scrape/munge public clarivate dumps
 - repositories (?)
diff --git a/notes/explore.md b/notes/explore.md
index 5f23d35..c25404d 100644
--- a/notes/explore.md
+++ b/notes/explore.md
@@ -12,6 +12,17 @@ PKP PLN numbers result in?
 
 So about 60k releases.
 
+How about Hathitrust?
+
+    select count(*), sum(journal.release_count), sum(journal.preserved_count) from journal join directory on journal.issnl = directory.issnl where directory.slug = 'hathitrust';
+
+    count(*)    sum(journal.release_count)  sum(journal.preserved_count)
+    ----------  --------------------------  ----------------------------
+    26628       48160184                    36905342
+
+Much larger potential impact, of 11+ million releases, though unclear how many
+are acutally in the hathitrust archives.
+
 ## 2020-06-23
 
 Where do back ISSN-Ls come from? Answer: exiting fatcat metadata.