Changes between Version 2 and Version 3 of Ticket #15844, comment 19


Ignore:
Timestamp:
May 18, 2015, 11:42:05 PM (4 years ago)
Author:
leeroy
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • Ticket #15844, comment 19

    v2 v3  
    1 I've been familiarizing myself with the code in an effort to understand the properties of the various datatypes. I would like to ensure I'm correct in my understanding: Bridges can only be located by sha1 of fingerprint? Relays can be located by fingerprint, sha1 of fingerprint, or base64 encoding of fingerprint?
     1I've been familiarizing myself with the code in an effort to understand the properties of the various datatypes. I would like to ensure I'm correct in my understanding: Bridges can only be located by sha1 of fingerprint? Relays can be located by fingerprint, sha1 of fingerprint, or base64 encoding of fingerprint? A function index using digest() and encode() will ensure all three are fast with the advantage of only storing one.
    22
    3 A function index for relays may come in handy here. The fingerprint search needs to be fast irregardless. So store the fingerprint, create a function index using digest() and encode(). Now all three are fast with the advantage of only storing one.
    4 
    5 I would also like to ask how strict is the protocol wrt the ordering of keys in JSON responses? A client should be able to render the view irregardless, but how true is this in practice? I wouldn't want to break anything when using pgnosql.
     3I would also like to ask how strict is the protocol wrt the ordering of keys in JSON responses? For example using Atlas as a client. I wouldn't want to break anything when using pgnosql. The RFC defines the set as unordered for interoperability.