aboutsummaryrefslogtreecommitdiff
path: root/proposals/276-lower-bw-granularity.txt
diff options
context:
space:
mode:
authorNick Mathewson <nickm@torproject.org>2017-02-26 19:35:32 -0500
committerNick Mathewson <nickm@torproject.org>2017-02-26 19:35:32 -0500
commitb1e94492ac661db37fab2efc0182d2b2998d5e3b (patch)
tree2dd927634005f5d4f63588770b87b8c6b65be5ae /proposals/276-lower-bw-granularity.txt
parent2e5e0cb3f87f6813b789f09459daea6ebcaa4eb4 (diff)
downloadtorspec-b1e94492ac661db37fab2efc0182d2b2998d5e3b.tar.gz
torspec-b1e94492ac661db37fab2efc0182d2b2998d5e3b.zip
276 update w comments from teor
Diffstat (limited to 'proposals/276-lower-bw-granularity.txt')
-rw-r--r--proposals/276-lower-bw-granularity.txt16
1 files changed, 10 insertions, 6 deletions
diff --git a/proposals/276-lower-bw-granularity.txt b/proposals/276-lower-bw-granularity.txt
index 4d3735c..bca544b 100644
--- a/proposals/276-lower-bw-granularity.txt
+++ b/proposals/276-lower-bw-granularity.txt
@@ -33,20 +33,24 @@ Target: 0.3.1.x-alpha
3. Proposal
- I propose that we round the bandwidth values as they are placed in
- the votes to two no more than significant digits. In addition, for
+ I propose that we round the bandwidth values, as they are placed in votes,
+ to no more than two significant digits. In addition, for
values beginning with decimal "2" through "4", we should round the
first two digits the nearest multiple of 2. For values beginning
with decimal "5" though "9", we should round to the nearest multiple
of 5.
- This change does not require a consensus method; it will take effect
- once enough authorities have upgraded.
+ The change will take effect progressively as authorities upgrade: since
+ the median value is used, when one authority upgrades, 1/5 of the
+ bandwidths will be rounded (on average).
+
+ Once all authorities upgrade, all bandwidths will be rounded like this.
4. Analysis
- The rounding proposed above will not round any value by more than
- 5%, so the overall impact on bandwidth balancing should be small.
+ The rounding proposed above will not round any value by more than 5% more
+ than current rounding, so the overall impact on bandwidth balancing should
+ be small.
In order to assess the bandwidth savings of this approach, I
smoothed the January 2017 consensus documents' Bandwidth fields,