- Issue created by @larowlan
- Status changed to Postponed
2 months ago 5:12am 4 February 2025 - π¦πΊAustralia larowlan π¦πΊπ.au GMT+10
I think this needs π Support server side massage and validation of component prop form values Active
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
The intent here is purely a UX improvement, right?
i.e.:
- eliminate one request
- β¦ a request that can only be made after another request's response has been processed by the client
- All to improve the TTI, right?
If so: why is this blocked on π Support server side massage and validation of component prop form values Active ? What am I missing?
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
Having reviewed #3499550 in some detail (see [##3499550-13]) AFAICT this should land first and will then make #3499550's MR simpler.
- π¦πΊAustralia larowlan π¦πΊπ.au GMT+10
The intent is partially performance but also it gets us closer to π META: Conflict free concurrent editing Active
We need to do π Support server side massage and validation of component prop form values Active first because it introduces the previewSlice
- π¦πΊAustralia larowlan π¦πΊπ.au GMT+10
Commit ace768f6 has the work here on top of π Support server side massage and validation of component prop form values Active
When that is in, we can rebase --onto origin/0.x and drop d78ade44
- π¦πΊAustralia larowlan π¦πΊπ.au GMT+10
Also blocked on π OpenAPI spec insufficiently precise for `LayoutComponent` Active which will make the diff here smaller
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
π Support server side massage and validation of component prop form values Active is in!
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
π OpenAPI spec insufficiently precise for `LayoutComponent` Active also landed.
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
- π¦πΊAustralia larowlan π¦πΊπ.au GMT+10
@heyyo mentioned this in π Improve backend API when lots of components Active
From slack
It looks like a POST request, why we need to post anyhting to backend when just loading the page ?
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
+1 β this seems like super low-hanging fruit!
- Merge request !820Issue #3492061 - Add the preview HTML to ApiLayoutController:get β (Merged) created by larowlan
- π¦πΊAustralia larowlan π¦πΊπ.au GMT+10
7 e2e fails feels like real things
- π¬π§United Kingdom longwave UK
Added some nits but as usual this looks good, leaving at NR so Wim or others can still review.
- First commit to issue fork.
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
So much awesome here β I did not expect this to help harden our end-to-end test infrastructure, but it's great to see that happen!
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
Looks splendid now, and apparently my request for clarification was raised by @larowlan ~2 months ago too, so he did it in this MR, which enabled him to close π [PP-1] Consider renaming 'initialized' boolean in layout slice to 'postPreview'' Postponed π
This was only missing @bnjmnm's explicit approval of the MR for
ui/src/components/form/inputBehaviors.tsx
for+ // Flag that we need to update the preview. + dispatch(setUpdatePreview(true)); dispatch(setPageData(values));
This seems low-risk enough to just go ahead and merge this, especially considering @bnjmnm AFAICT already reviewed this MR (see #20).
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
π Get Options as buttons in Page Data form working Active seemed to unblock more things, so I landed that first. It conflicted with this. I tried to resolve it. Let's see what CI says.
-
wim leers β
committed d85bf556 on 0.x authored by
larowlan β
Issue #3492061 by larowlan, wim leers, bnjmnm, longwave: Include the...
-
wim leers β
committed d85bf556 on 0.x authored by
larowlan β
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
This unblocked π Replace the postPreview action with atomic equivalents Active ! π