-
Notifications
You must be signed in to change notification settings - Fork 4
Automatic reloading of results does not work as expected #225
Comments
I removed the hard-coded minimum number of initial rows. The server-side config will always be used. |
There are still some issues when automatically loading further results.
|
This has not happened to me. Is it reproducable?
The displayed number is the number of all available results which can be loaded by lazy loading, not the number of loaded results.
Are you sure you are using the respective Rocket.Chat branch with the changes? (smarti-widget-0.7.1) |
All this can be reproduced on out test env where you have access to.
Yes, I understand. But when 71 results are possible why does the lazy loading stop at round about 10?
Yes I am. Have a look at our test env: |
Which browsers did you test? |
I checked the code on the test env and I couldn't find my rocket.chat changes. |
We have deployed the Smarti artifact from https://github.com/redlink-gmbh/smarti/releases/tag/smarti-0.7.1 So this is independent from RC version |
And regarding the 0 results: When the displayed number is not 0 and no more results are loaded, it's most probably because of a horizontal scrollbar on the bottom which prevents lazy loading. That's a UI bug and I will fix it. |
* Smarti-203: Flag widget messages * Smarti-203: Rename widget message flag to skipAnalysis * Smarti-52: Add support for edited messages * redlink-gmbh/smarti#221: Set room topic as support area for conversation * Revert "redlink-gmbh/smarti#221: Set room topic as support area for conversation" This reverts commit c7966a3. * redlink-gmbh/smarti#225: Fix widget footer styling * redlink-gmbh/smarti#220: Improve error handling * SmartiWidgetBackend: Improve error handling
I'm working on an improved error handling in the widget. |
The problems regarding the lazy loading behavior and the displayed result count are still persistent.
@Peym4n Can you please verify, that those versions include you expected changes? |
Yes, the changes are correctly deployed. |
Referring to this comment:
See #255
See #256 Closing in favor of new issues. |
I would like to let this issue opened until we have tested the upcoming release. |
Moved the issue in: Needs Review |
@ruKurz ok. makes sense. |
(cherry picked from commit 406eca9)
There are still some issues when automatically loading further results.
Consider open points, that are not yet tested as done within this issue:
Those will be continued here: |
* Fixing #345 - where outdated conversation id could not be found in smarti * Made serveral changes regarding the review. Added an admin option to re-synch completely. * Remove conversations from mapping cache, if a room is being ereased. Ordered private dunctions and added some comments. * Fixed i18n keys. * Get analysis and notify room after deleting a message. * Removed rate limiter and explicitly perform an conversation analysis. * Fixed several synchronisation issues related to the LivechatExternalMessage cache. * Removed interface method getRoomId. * removed unused SystemLogger * Improved error handling and debug/error messages. * Added missing _ definition * Resynch Limiter optimizied Restricted resynch access for admins Extracted synch to Smartiadapter Using getProperties instead legacy impl when sync * Refactored SmartiRelaod * Moved hooks to Assistify AI. Simplified clear method. * Make sure conversations being created when a new request, topic or thread is created * fixed system messages in smarti conversations * better rely on cache as long as the "legacy rc endpoint" does not work. * - Extracted functions - Increased readability - Code cleanups - Added comments * Fixing undefined vars & trailing-spaces * fixing undefined rid * Don't flush the cache if legacy webservice is not used * auto resync on message, resync livechat rooms * redlink-gmbh/smarti#225: Improve error handling (cherry picked from commit 406eca9) * fix duplicated room creation on new room * fixed test and minor bug in smarti adapter * fix linting * Use leagcy rocket.chat endpoint again when getting conversations * fixed assistify tests * removed redundant error logging from proxy * Improved health check before resynch * fix tests
Expected Behaviour
After scrolling down the search results, next results should be loaded dynamically.
Actual Behaviour
In some cases the lazy loading stops, even there are results left.
Steps to Reproduce
Observations
numfound=13
but a Solr query withstart=16
is executedrows=6
) on the client side by default to completely fill the hit display area. However, the hits per page are configured on the server side by design, whereby a separate value can be set for each Smarti client configuration. Therefore, the value for rows is always overwritten by the server-side configuration. (BTW: To fill the area of the result display completely, the available space for the hit display and the size of the individual results should be calculated/considered.Acceptance criteria
Environment
0.7.0
1.8
The text was updated successfully, but these errors were encountered: