Storage "Interface Driven Design" Document#761
Conversation
|
UPDATE: I have added everything except ACL sections. What a doozy! Please start reviewing. (I'm going to hold off on cc'ing Travis until I finish the ACL part.) It's encouraged me to just throw away a few methods (e.g. |
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
|
@tseaver All the comments without response I plan on addressing. I am away from keyboard today and will get to them tomorrow (March 28, 2015). |
|
@tseaver I have added ACL stuff and changed around a few of the things you already reviewed. |
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
d84dc38 to
ca5122d
Compare
Still in flux (as of filing).
Follow along changes in:
www.googledrive.com/host/0B8el7dRo8mVOMk1HbFk4TVhDaEE
/cc @jgeewax @tseaver I'll ping Travis when it is complete.
UPDATE (March 30): @thobrla PTAL. @jgeewax asked me to design the final state of the desired surface area (all inspired by your bug #545). Our current code uses
projection=noAclto avoid loading ACL info with other metadata, while this hypothetical design ditches that concept.