Using Twitter as one of our annotations repositories?

would it make sense to exploit twitter
as a general literary-annotations database?

it appears twitter will index hashtags of any length
so we should be able to
compose
unique hashtags
associated with any web document
at any degree of granularity
and use those hashtags
to track annotations of these documents
over a seemingly unlimited time frame

so
#u003called
could be a reasonably unique hashtag
for debate of the "called up"/"called out" question
('u' for ulysses
'003' for page 3 of the 1922 edition
'called' as the least ambiguous single keyword)

similarly
#fw003riverrun
for finnegans wake
and maybe even
#fw001apostrophe
for finnegans vs finnegan's

if we post every current fweet note
as a separate tweet
that's currently under 100,000
(there's about 200,000 words,
1.3M characters)

even at the snailpace of four tweets per hour
that would only take three years
(but does twitter even throttle this at all?)

and if we're planning ahead for all documents
"u" and "fw" will be inadequate
but "jaju" and "jajfw" much less so
(more important docs deserve the shortest hashkeys)

fweet includes lots of crosslinks for motifs, etc
eg #fwcoleridge
#fwfrench

some intentional spam might need to be blocked

enthusiasts can follow each others' feeds
and use the hashtags to call up past debates
and discover new contributors

ambiguous spellings can be tagged both ways

phrases can be tagged for each important word:
#fw003past #fw003eve #fw003adams
(punctuation may need special handling)

including a fweet or pjoyce url will still leave 120chars
minus at most 20 per hashtag

long annotations can be split up or handled via a url
(ellipses could signal
twitter-threaded-replies
without hashtags)

if twitter supports full unicode in hashtags
unique hashtags can be a lot shorter
(though you'll need to copy and paste them rather than typing them in)

maybe whole chapters can get tags:
#u003telemachus

images, videos and links can be included as annotations