Open bryanhuntesl-lowpriv opened 7 years ago
Thoughts from meeting on 2017-10-06:
General directive to Development Group's Teams is to:
Just purning branches can be complicated i think we need to spend some time in identifying what is build from what before we go berserk. I (think) that TS is build from a different riak_core branch then KV for example. Ideally we'd eventually get away with all of that (please!) but if we just delete everything we don't recognize it could be quite desasterous.
Agreed. Motivation is to reduce audit area.
On Sat, 7 Oct 2017 03:04 Heinz N. Gies, notifications@github.com wrote:
Just purning branches can be complicated i think we need to spend some time in identifying what is build from what before we go berserk. I (think) that TS is build from a different riak_core branch then KV for example. Ideally we'd eventually get away with all of that (please!) but if we just delete everything we don't recognize it could be quite desasterous.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/TI-Tokyo/riak-community-admin/issues/19#issuecomment-334903539, or mute the thread https://github.com/notifications/unsubscribe-auth/AfDuAzdOWoTDDmndJVe0BWu93H94VBWpks5sptwagaJpZM4PwfW1 .
Basho repositories contain enormous number of unmerged branches. I am also unsure as to what constitutes ‘develop’ or what branching policy (if any) basho were using.
I wrote list_big_branches.py to investigate.
I list below the list top 30 repositories by number of open branches.
This is clearly ridiculous.
riak_kv,462
riak_core,272
riak_test,212
riak_cs,195
riak,123
leveldb,105
eleveldb,101
basho_bench,89
yokozuna,89
erlang-pbkdf2,76
bitcask,74
riak_ensemble,65
riak-java-client,61
riak_api,61
machi,58
riak_ql,51
riak_search,50
riak_dt,49
riak-erlang-client,45 riak_pipe,44
riak_control,44
riak_pb,43
stanchion,37
CI_test,35
basho_docs,32
lager,28
node_package,27
webmachine,27
s3-tests,26
cuttlefish,26