Commit Graph

3508 Commits

Author SHA1 Message Date
Jason Zimdars 23bd8a569e This is annoying 2025-04-23 08:53:43 -05:00
Jorge Manrubia 970a51db28 No need for it! 2025-04-23 15:49:22 +02:00
Jorge Manrubia b4cfd40a2f Inline, no need to duplicate the conditional 2025-04-23 15:37:28 +02:00
Jorge Manrubia 6f98f2f573 Order methods 2025-04-23 15:32:22 +02:00
Jorge Manrubia e964c5f9c9 Not used! 2025-04-23 15:31:26 +02:00
Jorge Manrubia 53eba9bd0b Favor after_create_commit to keep transactions fast 2025-04-23 15:28:28 +02:00
Jorge Manrubia ea66845e71 There is always an action 2025-04-23 15:27:58 +02:00
Jorge Manrubia e8e95bd511 Use it 2025-04-23 15:27:25 +02:00
Jorge Manrubia 4bef466df2 1-line 2025-04-23 15:22:41 +02:00
Jorge Manrubia ad58de9e76 Remove access checks when watching
https://github.com/basecamp/fizzy/pull/425#discussion_r2056039581
2025-04-23 15:22:18 +02:00
Jorge Manrubia a9df2d48d1 Remove delegate 2025-04-23 15:17:51 +02:00
Jorge Manrubia 6ff9f7a154 Fix identation 2025-04-23 15:07:02 +02:00
Jorge Manrubia 469fb9f049 Simplify the logic to determine action text changes 2025-04-23 15:02:31 +02:00
Jorge Manrubia e446080367 Don't create duplicated mentions
E.g: when editing a comment or description with a mention
2025-04-23 14:45:56 +02:00
Jorge Manrubia f43629f849 Format 2025-04-23 14:39:10 +02:00
Jorge Manrubia dc5f64ab30 Remove redundant default queue declaration 2025-04-23 14:37:09 +02:00
Jorge Manrubia ab73417e73 Leverage existing resolve route for comments 2025-04-23 14:34:59 +02:00
Jorge Manrubia 07d3447e7e Resolve notifications URLs 2025-04-23 14:34:59 +02:00
Jorge Manrubia 2eae58725a Remove notifications' resource relationship
Is confusing since "source" already captures the origin of the notifications. It's
needed to interpret things at rendering time, so we can query things as needed there.
2025-04-23 13:52:47 +02:00
Jorge Manrubia 696732a90a Rename mentions' container => source 2025-04-23 13:41:34 +02:00
Jorge Manrubia 7c0aff0680 Don't indent case 2025-04-23 13:36:39 +02:00
Jorge Manrubia 043230394e Tidy up event helper code
Also, revert to using a helper method to determine the event action, since this is purely a view concern. Keep method at the event model to identify initial assignments.
2025-04-23 13:35:09 +02:00
Jorge Manrubia 2e3ffcfb9d Remove identation 2025-04-23 13:35:08 +02:00
Jorge Manrubia 755c9b45c7 Rename and fix problem where after_touch was overriding the saved content 2025-04-23 12:38:33 +02:00
Jorge Manrubia edcc9e3279 Add mentionees as watchers 2025-04-23 12:38:06 +02:00
Jorge Manrubia 8ebb3f2fa9 We need to remember on after_touch so that the mechanism triggers when touch-cascading records automatically via associations 2025-04-23 11:38:34 +02:00
Jorge Manrubia 8b6cfccded Rename to clarify usage 2025-04-23 11:20:33 +02:00
Jorge Manrubia ab54cafd84 Inline event notifiers 2025-04-23 11:06:43 +02:00
Jorge Manrubia b510ca0318 Don't choke for nil values 2025-04-23 10:56:00 +02:00
Jorge Manrubia 68e2e96076 Fix indentation 2025-04-23 10:44:42 +02:00
Jorge Manrubia af8701567a Extract method 2025-04-23 10:35:18 +02:00
Jorge Manrubia c3cd873588 Rename collect => create 2025-04-23 10:30:30 +02:00
Jason Zimdars 14d0b995a8 Fix summary 2025-04-22 23:09:29 -05:00
Jason Zimdars 8fec68c34c Remove index columns 2025-04-22 23:07:06 -05:00
Jason Zimdars b29afcd518 Same treatment for the events index 2025-04-22 22:57:48 -05:00
Jason Zimdars 0a9d16bf2d New menu filters and navigates 2025-04-22 22:29:33 -05:00
Jason Zimdars 72c1fa3b05 Merge branch 'main' into fixes-tues-22
* main:
  Indention
  @collection is always available when card scoped
  Style
  Duh, no before destroy commit, but this is still not right!
  No long transactions!
  But feels more essentially part of the User, so list first
  Not related to the other two
  Style
  Try a slimmer version
  We werent using any of the other avatar methods
  Dont do a lot of work inside a single transaction
  Really liking the new it-style yields in many places
  Bump nokogiri from 1.18.7 to 1.18.8
2025-04-22 15:06:36 -05:00
Jason Zimdars 537903ba0b Don't test for _Added by_ events
Note: The join method in `event_summary.rb` inserts a space that we need to account for
2025-04-22 15:06:05 -05:00
David Heinemeier Hansson cf3cbccf23 Indention 2025-04-22 21:56:23 +02:00
David Heinemeier Hansson 93516f2e48 @collection is always available when card scoped 2025-04-22 21:51:38 +02:00
David Heinemeier Hansson dff55d6cb6 Style 2025-04-22 21:48:32 +02:00
Jason Zimdars 647971f205 Remove check for _Added by..._ line 2025-04-22 14:46:46 -05:00
Jason Zimdars 1412932666 The placeholder attribute is on the house-md element 2025-04-22 14:45:39 -05:00
David Heinemeier Hansson 5048e9c970 Duh, no before destroy commit, but this is still not right! 2025-04-22 21:40:30 +02:00
David Heinemeier Hansson 7124835e60 No long transactions!
If you intend to deactivate someone, and the process fails mid process,
so you only delete some sessions, or some accesses, you are actually
fine. The system is never left in an incomplete state. And that's really
the only time we should be using transactions with sqlite3 -- to prevent
actual data integrity issues.

The rest of the time, we should optimize for each transaction to be as
quick as possible. The sqlite3 writer lock is the one bottleneck that we
cant afford to overload.
2025-04-22 21:29:03 +02:00
Jason Zimdars 5de85fa6bf Don't display published event in threads 2025-04-22 14:25:35 -05:00
David Heinemeier Hansson 90f05b7310 But feels more essentially part of the User, so list first 2025-04-22 21:25:30 +02:00
David Heinemeier Hansson 72d1d0baee Not related to the other two 2025-04-22 21:24:51 +02:00
Jason Zimdars 8c43615d71 Missing controller 2025-04-22 14:23:44 -05:00
David Heinemeier Hansson ad1e97767d Style 2025-04-22 21:22:42 +02:00