projects / mediagoblin.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 98853c6) | patch
Switch JSONEncoded from VARCHAR to TEXT
authorElrond <elrond+mediagoblin.org@samba-tng.org>
Sun, 1 Apr 2012 17:55:15 +0000 (19:55 +0200)
committerElrond <elrond+mediagoblin.org@samba-tng.org>
Sun, 1 Apr 2012 17:58:23 +0000 (19:58 +0200)
commit1e3a0e0ca0c58ae740424fa6cf239f2dba5dc592
tree124184737dfe2ce53866776893625b58371fba6etree | snapshot (zip tar.gz)
parent98853c63fde6e133dca6c723aaacfb1f368cad38commit | diff
Switch JSONEncoded from VARCHAR to TEXT

The JSON fields are really "dumb stuff in here" fields.
They are not intended to get indexed or anything. And they
can get large. For example the exif_all field in one of my
simple tests is nearly 7 kB large. Although VARCHAR might
work, TEXT feels just better as the storage type.
mediagoblin/db/sql/extratypes.py diff | blob | blame | history
Unnamed repository; edit this file 'description' to name the repository.