You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[$500] Workspace - Endless skeleton loader on the member page if you switch to it multiple times #40595
Version Number: 1.4.63.7 Reproducible in staging?: Y Reproducible in production?: Y If this was caught during regression testing, add the test name, ID and link from TestRail: N/A Issue reported by: Applause - Internal Team
Action Performed:
Log in with a new expensifail account
Create a workspace
Navigate to the members page
Navigate back
Navigate to the members page again and wait 20-30 seconds
Navigate back
Navigate to the members page again
Expected Result:
Skeleton loader should only be visible for a very short time
Actual Result:
Endless skeleton loader appears on the member page if you switch to it multiple times
Workaround:
Unknown
Platforms:
Which of our officially supported platforms is this issue occurring on?
👋 Friendly reminder that deploy blockers are time-sensitive ⏱ issues! Check out the open `StagingDeployCash` deploy checklist to see the list of PRs included in this release, then work quickly to do one of the following:
Identify the pull request that introduced this issue and revert it.
@puneetlath FYI I haven't added the External label as I wasn't 100% sure about this issue. Please take a look and add the label if you agree it's a bug and can be handled by external contributors
changed the title [-]Workspace - Endless skeleton loader on the member page if you switch to it multiple times[/-][+][$250] Workspace - Endless skeleton loader on the member page if you switch to it multiple times[/+]on Apr 20, 2024
Upon thorough analysis, it's evident that the endless skeleton loader arises from delays in data fetching and rendering when switching to the member page multiple times. To address this issue, I propose the following technical changes:
I'll implement efficient data fetching strategies aimed at reducing latency and improving loading times. This may involve optimizing network requests, employing caching mechanisms, or exploring asynchronous data loading techniques.
By revisiting the component rendering logic, I'll prioritize the display of essential content while asynchronously loading additional data in the background. This approach will mitigate the appearance of the endless skeleton loader and provide users with meaningful content promptly.
I'll introduce robust error handling mechanisms to gracefully manage data retrieval failures and network interruptions. Users will be presented with informative error messages, and detailed error logs will be generated for debugging purposes.
Integration of performance monitoring tools will be crucial for tracking page load times and identifying potential performance bottlenecks. Continuous monitoring will enable me to refine the member page loading experience and maintain optimal performance levels.
I hope this solution will help you with this issue.
and since the workspace member list cannot be empty (at least the owner is in the member list), we can show the message Cannot display the member list right now! Please try again! (this is subject to change) by modifying the getHeaderMessage function
const getHeaderMessage = () => {
if (isOfflineAndNoMemberDataAvailable) {
return translate('workspace.common.mustBeOnlineToViewMembers');
}
return isEmptyObject(policy?.employeeList) ? 'Cannot display the member list right now! Please try again!': '';
};
Still can not reproduce this issue but I don't agree with you at this point, employeeList can be empty during it's fetching from server for the first time user open the app.
Since this issue was not reproduced anymore, I just wonder if we can close it?
@hungvu193 the empty array value of employeeList comes from the OpenApp API, the mergeCollection for the policy_ collection operation. It's a separate issue and the BE should take a look into this?
However, if the OpenApp returns, the employeeList again cannot be empty, or at least it cannot be null unless the situation described in my proposal happens, thus my assumption is not totally inaccurate
the empty array value of employeeList comes from the OpenApp API,
I think this is intentional. So I still don't agree with you at that point.
cc @puneetlath for your thoughts at above discussion, I think we can close this issue.
@hungvu193 to be honest, I don't think it should be intentional that way. The onyx updates coming from the OpenApp API call should be atomic, which means FE should only care about the final state after apply all the Onyx operations at once. So, under no circumstances should the policy?.employeeList be promptly [] before becoming non-empty again, because OpenApp API call is just 1 single operation, which return the policy data with non-empty employeeList (even though it contains multiple Onyx sub-operations). @puneetlath can we have someone on the BE team to check this point, as I believe we shouldn't split into 2 operations mergeCollection and merge for policy_XXX just to update the employeeList from [] to {[email]: {....}}. It should live either in mergeCollection operation.
Again, it still confirms that the employeeList, by definition, cannot be empty.
Activity
melvin-bot commentedon Apr 19, 2024
Triggered auto assignment to @puneetlath (
DeployBlockerCash), see https://stackoverflowteams.com/c/expensify/questions/9980/ for more details.github-actions commentedon Apr 19, 2024
👋 Friendly reminder that deploy blockers are time-sensitive ⏱ issues! Check out the open `StagingDeployCash` deploy checklist to see the list of PRs included in this release, then work quickly to do one of the following:
lanitochka17 commentedon Apr 19, 2024
@puneetlath FYI I haven't added the External label as I wasn't 100% sure about this issue. Please take a look and add the label if you agree it's a bug and can be handled by external contributors
lanitochka17 commentedon Apr 19, 2024
We think that this bug might be related to #wave-collect - Release 1
[-]Workspace - Endless skeleton loader on the member page if you switch to it multiple times[/-][+][$250] Workspace - Endless skeleton loader on the member page if you switch to it multiple times[/+]melvin-bot commentedon Apr 20, 2024
Job added to Upwork: https://www.upwork.com/jobs/~01e843a4238c37c0d9
melvin-bot commentedon Apr 20, 2024
Triggered auto assignment to Contributor-plus team member for initial proposal review - @hungvu193 (
External)digitalwaver commentedon Apr 20, 2024
Upon thorough analysis, it's evident that the endless skeleton loader arises from delays in data fetching and rendering when switching to the member page multiple times. To address this issue, I propose the following technical changes:
I hope this solution will help you with this issue.
Cooper Brown from Upwork.
36 remaining items
hungvu193 commentedon May 8, 2024
Still can not reproduce this issue but I don't agree with you at this point,
employeeListcan be empty during it's fetching from server for the first time user open the app.Since this issue was not reproduced anymore, I just wonder if we can close it?
dominictb commentedon May 8, 2024
@hungvu193 the empty array value of
employeeListcomes from the OpenApp API, themergeCollectionfor thepolicy_collection operation. It's a separate issue and the BE should take a look into this?However, if the OpenApp returns, the employeeList again cannot be empty, or at least it cannot be null unless the situation described in my proposal happens, thus my assumption is not totally inaccurate
hungvu193 commentedon May 8, 2024
I think this is intentional. So I still don't agree with you at that point.
cc @puneetlath for your thoughts at above discussion, I think we can close this issue.
melvin-bot commentedon May 8, 2024
📣 It's been a week! Do we have any satisfactory proposals yet? Do we need to adjust the bounty for this issue? 💸
dominictb commentedon May 8, 2024
@hungvu193 to be honest, I don't think it should be intentional that way. The onyx updates coming from the
OpenAppAPI call should be atomic, which means FE should only care about the final state after apply all the Onyx operations at once. So, under no circumstances should thepolicy?.employeeListbe promptly[]before becoming non-empty again, becauseOpenAppAPI call is just 1 single operation, which return thepolicydata with non-emptyemployeeList(even though it contains multipleOnyxsub-operations). @puneetlath can we have someone on the BE team to check this point, as I believe we shouldn't split into 2 operationsmergeCollectionandmergeforpolicy_XXXjust to update theemployeeListfrom[]to{[email]: {....}}. It should live either inmergeCollectionoperation.Again, it still confirms that the
employeeList, by definition, cannot be empty.mvtglobally commentedon May 13, 2024
Issue not reproducible during KI retests. (Second week)
melvin-bot commentedon May 13, 2024
@puneetlath, @hungvu193 Huh... This is 4 days overdue. Who can take care of this?
melvin-bot commentedon May 15, 2024
📣 It's been a week! Do we have any satisfactory proposals yet? Do we need to adjust the bounty for this issue? 💸
melvin-bot commentedon May 15, 2024
@puneetlath, @hungvu193 6 days overdue. This is scarier than being forced to listen to Vogon poetry!
puneetlath commentedon May 16, 2024
Ok, based on the discussion I agree that we can close this out. Thanks everyone.