Merge conflict Gerrit cannot merge this change. It needs a rebase onto main.

Review main Work in progress Orphaned

90171[BUGFIX] Allow to review workspace changes of all mounts or sites

Benjamin Franzke CR 0 V -1 Easy +13 −10 updated 07/30/25, 12:54

Commit message

(Re)allow workspace users to get an overview of all changes of all
mounts or sites by navigating to the virtual page uid=0.

This behaviour has previously been fixed, via #95854, regressed due
to #96548 and is now fixed by differentiating between an implicit
id=0 and an explicit id=0 selection by the user.

The explicit id=0 selection now renderes the workspace overview
of all sites and webmounts, whereas an context-free (no page selected)
initial module load will render the behaviour as established by #96548.

Test this change

Run a command in your clone of the TYPO3 Core. Then try what the change does and vote on Verified in Gerrit. How to review a patch

Check out patch set 1
git fetch https://review.typo3.org/Packages/TYPO3.CMS refs/changes/71/90171/1 && git checkout FETCH_HEAD
Cherry-pick onto main
git fetch https://review.typo3.org/Packages/TYPO3.CMS refs/changes/71/90171/1 && git cherry-pick FETCH_HEAD

Review

  • Code review CR 0 Needs a vote
  • Verified V -1 Blocked
    • Oli Bartsch -1
    • core-ci +1 CI

Issues 3

Files 2

Review log 910

  1. Benjamin Franzke uploaded patch set 1 07/28/25, 11:47

  2. core-ci voted on patch set 1 V +1 07/28/25, 11:54 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/82421

  3. Oli Bartsch commented on patch set 1 07/30/25, 08:54

    To be honest, I don’t think it’s a good idea to allow users to publish or discard entire workspaces from this view. One of our customers, recently migrated from in2publish to workspaces, accidentally published an entire workspace on the second day. This might be a UX issue — but if so, that should be addressed first. Also, I don’t see much benefit in listing all records here, especially since there could be a huge number. The "depth" selection doesn’t seem to work in this view either, which further limits usability.
  4. Benjamin Franzke commented on patch set 1 07/30/25, 09:21

    Reply

    Our customers (of the company that I worked for when I filed https://forge.typo3.org/issues/95854) really needed this (after it regressed it before), as referenced in the paragraph that follows this sentence. How yould you allow a reviewer of 50 sites get an overview over changes in all sites? (that was our usecase) Should that manager be required to iterate through all these 50 sites manually? > accidentally published an entire workspace on the second day with that argument you'd need to disallow any depth != `1`, you can always publish "too much", that shouldn't limit a view, but rather qualifies for usability improvements that warn a user if a lot of actions will be performed (like slack `@channel` notifications) > The "depth" selection doesn’t seem to work in this view either, which further limits usability. Please share what's not working, it's working for me. Also: Broken functionality would not be a blocker, but a reason to fix it.
  5. Oli Bartsch commented on patch set 1 V -1 07/30/25, 09:40

    Reply

    > How yould you allow a reviewer of 50 sites get an overview over changes in all sites? I don’t think reviewing changes across 50 sites simultaneously is a typical use case. How is a user supposed to meaningfully comprehend that volume of changes at once? I’m not opposed to this view in general. But in its current state, as outlined, it lacks important UX safeguards. There’s no additional confirmation when publishing or discarding changes across many sites, which increases the risk of unintended actions. We should IMO resolve those before enabling such view. The "depth" selection doesn’t seem to have any effect on the listing. It still shows all changes across all levels, regardless of the selected depth. This makes it harder to limit the scope, especially with 50 sites. I've added a video to the issue https://forge.typo3.org/attachments/39760. Side note: Given that "depth" works it is even more complicated for the user since the mass action is NOT limited to the listing but still published / discards ALL record of the workspace.
  6. Benjamin Franzke commented on patch set 1 07/30/25, 11:08

    Reply

    > it lacks important UX safeguards. There’s no additional confirmation when publishing or discarding changes across many sites, It's the same risk that we currently already have when recursively publishing content of one site. TBH: It feels like (I know you said you're not opposed in general) you want to block this regressionfix via new requirements and side quests. I disagree on that approach for regressions – I understand that for new FEATURES, but not for regressions – but I'll of course accept your vote. I'll come back to this patch and provide pre-patches to address the side quests, since I don't have time to work on this at the moment. Which simply means our users will need to wait longer 😞 => Marking as WIP
  7. Benjamin Franzke marked it work in progress 07/30/25, 11:08

  8. Oli Bartsch commented on patch set 1 07/30/25, 11:30

    Reply

    I appreciate your work on addressing this regression, and I understand the frustration when fixes get delayed. However, your last comment in my opinion tries to blame my honest review by stating that users need to wait longer now. Let me clarify: The "depth" functionality issue isn't a "side quest". It's a bug in the current patch that I've even provided a video for. I don't think fixing regressions by knowingly adding new bugs is our standard. You even mentioned yourself that "Broken functionality would not be a blocker, but a reason to fix it". I agree. That's exactly why we should fix this depth bug before merging. This has nothing to do with "side quests". Regarding the UX safeguards: I'm open to discussing the scope and necessity of additional confirmations prior to this patch. I'm not trying to block this regression fix. I'm trying to ensure we don't introduce new bugs while fixing old ones. Given our known high standards for patch quality, I'm confident you'll agree that shipping known broken functionality isn't the right approach, regardless of whether it's a regression fix or new feature. So, would you be able to address the depth functionality first? Once that works, we can have a look, if additional UX improvements are needed.
  9. Benjamin Franzke commented on patch set 1 07/30/25, 12:32

    Reply

    > The "depth" selection doesn’t seem to have any effect on the listing. It still shows all changes across all levels, regardless of the selected depth Thanks. Exactly this case works for me in pid=0 listings – depth 1 shwows only direct pages of pid=0 here, no subpages. I guess that's more than just a pid=0 bug that's exposed by your setup. > However, your last comment in my opinion tries to blame my honest review by stating that users need to wait longer now. I think you miss one point: We already had this functionality in the state you are raising as a blocker. It worked in TYPO3 v7, v8, v9, but broke in v10 for editors (because pid=0 was removed from the pagetree). During v12 development phase it was fixed for v12, v11 and v10 Later during v12 it broke again for v12, which means it kept broken for v13. That means: It worked in v7, v8, v9, v10, v11 with the exact state you are rasing as a blocker. It does not work in v12, v13 though, due to a cleanup patch. I want to fix it for v12 and v13 but we're now debating about UX issues that we have in any case and a bug that I – as written above – can not reproduce . It has usability problems? Ok, but that's not a reason to block readding functionality that was taken away. I understand, support and share your desire for good quality, but I think we need to differentiate between blockers for NEW functionality and blockers for REGRESSED functionality that we've taken away by accident or by an oversight. It's not blaming, but I make you responsible and want you to be aware of the implications. It's ok if we delay stuff if we really see the need. You do that in this case, I do not. I'm not happy that we need to disagree here, but I will diferently respect that. So, no, I'm not blaming you, but you still are responsible for the blocker. If we're blocking things because we see the need, that's fine and part of our job, but we take responsibility, which is fine. Thank you!
  10. Oli Bartsch commented on patch set 1 07/30/25, 12:54

    Reply Unresolved

    I need to correct a misunderstanding here: I am not blocking this patch - the bug is. You keep framing this as if I'm personally responsible for delaying users, but that's not correct. I documented a concrete bug and the responsibility lies with the patch author, not the reviewer. The argument that "it worked in v7, v8, v9 with the exact state you are raising as a blocker" doesn't justify knowingly reintroducing broken functionality. If anything, it highlights why we need proper quality assurance now. Instead of discussing "works on my setup / doesn't work on your setup", let's add regression tests, which should have been added anyway to ensure that the feature which worked in v7, v8, v9 will from now on work in v14, v15, and v20. This would prevent both the original regression AND the current bug from happening again.