Support blind key rotation #399

Merged

Plume_migration_agent merged 3 commits from blind-key-rotation into master

2019-01-05 21:30:29 +00:00

trinity-1686a commented

2018-12-30 11:30:10 +00:00

Owner

Fix #398

try to fetch user when receiving an invalid signature
regenerate new key-pair when sending Delete activity

Fix #398 - [x] try to fetch user when receiving an invalid signature - [x] regenerate new key-pair when sending `Delete` activity

codecov[bot] commented

2018-12-30 12:04:03 +00:00

(Migrated from github.com)

Codecov Report

Merging #399 into master will increase coverage by 0.21%.
The diff coverage is 0%.

@@            Coverage Diff             @@
##           master     #399      +/-   ##
==========================================
+ Coverage   27.87%   28.08%   +0.21%     
==========================================
  Files          63       63              
  Lines        7254     7405     +151     
==========================================
+ Hits         2022     2080      +58     
- Misses       5232     5325      +93

# [Codecov](https://codecov.io/gh/Plume-org/Plume/pull/399?src=pr&el=h1) Report > Merging [#399](https://codecov.io/gh/Plume-org/Plume/pull/399?src=pr&el=desc) into [master](https://codecov.io/gh/Plume-org/Plume/commit/3128e6a3b9963bda81f482e972eb853c0d564d35?src=pr&el=desc) will **increase** coverage by `0.21%`. > The diff coverage is `0%`. ```diff @@ Coverage Diff @@ ## master #399 +/- ## ========================================== + Coverage 27.87% 28.08% +0.21% ========================================== Files 63 63 Lines 7254 7405 +151 ========================================== + Hits 2022 2080 +58 - Misses 5232 5325 +93 ```

igalic (Migrated from github.com) reviewed 2018-12-30 14:28:28 +00:00

igalic (Migrated from github.com) left a comment

👀

src/routes/user.rs

igalic (Migrated from github.com) commented

2018-12-30 14:25:49 +00:00

should we be printing stuff here?

trinity-1686a reviewed 2018-12-30 14:38:00 +00:00

src/routes/user.rs

trinity-1686a commented

2018-12-30 14:38:00 +00:00

Author

Owner

I kept it because it was here before. If we had a proper logger this should get logged as this could be a an attack trial, but as it is, lost in stdout, I guess it's more of a debugging print?

igalic (Migrated from github.com) reviewed 2018-12-30 14:41:59 +00:00

src/routes/user.rs

igalic (Migrated from github.com) commented

2018-12-30 14:41:58 +00:00

i was wondering where our (debugging) log was

elegaanz (Migrated from github.com) reviewed 2019-01-04 19:20:03 +00:00

elegaanz (Migrated from github.com) left a comment

The code looks right, but I think I found a bug (maybe it's only me). To reproduce:

Create a@plume.one and b@plume.two
Make them follow each other
a@plume.one posts two articles, they are received on plume.two as they should
a@plume.one deletes one of these articles, and it is deleted on plume.two too
a@plume.one waits more than 10 minutes, and delete the second article
the Delete activity gets rejected by plume.two

Edit: also note that the next activities from a@plume.one are correctly received by plume.two

The code looks right, but I think I found a bug (maybe it's only me). To reproduce: - Create a@plume.one and b@plume.two - Make them follow each other - a@plume.one posts two articles, they are received on plume.two as they should - a@plume.one deletes one of these articles, and it is deleted on plume.two too - a@plume.one waits more than 10 minutes, and delete the second article - the `Delete` activity gets rejected by plume.two Edit: also note that the next activities from a@plume.one are correctly received by plume.two

elegaanz (Migrated from github.com) approved these changes 2019-01-05 21:16:32 +00:00

elegaanz (Migrated from github.com) left a comment

It is working now. 👍 (but I don't understand what was wrong with the previous condition, and this one doesn't make sense for me)

It is working now. :+1: (but I don't understand what was wrong with the previous condition, and this one doesn't make sense for me)

trinity-1686a commented

2019-01-05 21:27:51 +00:00

Author

Owner

previously, the first if would match in case of invalid request, and the second would do exactly the same, match on invalid request. But the first block is the Ok(()) one, so on invalid request it would say "ok this is fine".
Now the condition for the second if is inverted, so when the request is valid it returns Ok(()), when the request is invalid it returns the signature error

previously, the first `if` would match in case of invalid request, and the second would do exactly the same, match on invalid request. But the first block is the Ok(()) one, so on _invalid_ request it would say "ok this is fine". Now the condition for the second `if` is inverted, so when the request is _valid_ it returns Ok(()), when the request is invalid it returns the signature error

👍 2