Use `add_lease` instead of `renew_lease`
Fixes #220 (closed)
Maybe also #224 (closed)
Merge request reports
Activity
Created by: codecov[bot]
Codecov Report
Merging #226 (be4d9cf9) into main (e4ba6c78) will increase coverage by
0.01%
. The diff coverage is100.00%
.@@ Coverage Diff @@ ## main #226 +/- ## ========================================== + Coverage 96.97% 96.98% +0.01% ========================================== Files 49 49 Lines 3997 4011 +14 Branches 506 509 +3 ========================================== + Hits 3876 3890 +14 Misses 79 79 Partials 42 42
Impacted Files Coverage Δ src/_zkapauthorizer/lease_maintenance.py 95.38% <100.00%> (+0.07%)
src/_zkapauthorizer/tests/strategies.py 96.27% <100.00%> (ø)
...rc/_zkapauthorizer/tests/test_lease_maintenance.py 98.85% <100.00%> (+0.08%)
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update e4ba6c7...be4d9cf. Read the comment docs.Created by: tomprince
Review: Approved
This looks good to me.
I noticed that you didn't do the second part of my suggestion from #220 (closed). Looking at the code again, I guess this is because we don't have enough information (i.e. the cancel secret) to do so. I wonder if it makes sense to instead remove the method from
ZKAPAuthorizerStorageServer
and have it raiseNotImplemented
inZKAPAuthorizerStorageClient
(I guess the first one would have compatibility implications for our unreleased software)?This looks good to me.
I noticed that you didn't do the second part of my suggestion from #220 (closed). Looking at the code again, I guess this is because we don't have enough information (i.e. the cancel secret) to do so. I wonder if it makes sense to instead remove the method from
ZKAPAuthorizerStorageServer
and have it raiseNotImplemented
inZKAPAuthorizerStorageClient
(I guess the first one would have compatibility implications for our unreleased software)?Indeed. Thank you for following up on this point. I think maybe at the time I just forgot about that idea. Thinking about it now, esp. in the light of the fact that (a) Itamar discovered that there aren't actually any callers of
renew_lease
in the Tahoe-LAFS code base (apart from tests) and (b) Tahoe-LAFS has now actually droppedrenew_lease
from the network protocol entirely (ie you're not even allowed to call it anymore)...- Against a Tahoe-LAFS <1.16 storage server a custom client could call our
renew_lease
so if we provide it we still need to do ZKAP authorization there. But the standard client never does so we could also just delete ourrenew_lease
implementation entirely. - Against a Tahoe-LAFS >=1.16 storage server a custom client can no longer even get a
renew_lease
call deserialized (or whatever ... Foolscap will refuse to dispatch it) on the server. So for this case we could also just delete ourrenew_lease
implementation entirely. - No version of the Tahoe-LAFS client ever called
renew_lease
so the client-side version doesn't matter.
So ... I'll file another ticket for deleting all of our
renew_lease
code, I think, and apparently there aren't even any compatibility implications to worry about (though I guess I'll still worry about them and test them out somehow to be sure).- Against a Tahoe-LAFS <1.16 storage server a custom client could call our