-
Notifications
You must be signed in to change notification settings - Fork 8.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
"Create index pattern" wizard. #13154
"Create index pattern" wizard. #13154
Conversation
See additional comments: #12689 (comment) |
faf720c
to
b249ea1
Compare
For posterity, here is the previous |
@cjcenizal @chrisronline I wanted to use the current build to provide feedback and see if we can incorporate any of @alt74's designs. The PR doesn't seem to change anything in master at the moment. The create index pattern management screen looks the same to me. Is it ready to pull down? |
- Create a directive for each step in the wizard. - Reorganize files into conventional folder structure. - Rename files with more conventional and consistent naming patterns. - Display indices, partial matches, exact matches. - Add loading, empty, and success states. - Add option to include system indices.
b249ea1
to
e273403
Compare
@alexfrancoeur I just rebased against |
@chrisronline some initial feedback with the current PR There seems to be a refresh of the entire table as I type for an index pattern. Is there any way we can dynamically filter the results in the index table without doing a complete refresh? It'd be great to show where the pattern matches the index. Any chance we can bold where it matches? I'm struggling with how this can best work. If you look at the below screenshot, we are saying that an index pattern matches in it's current state. While the text matches the indices, the index pattern won't actually match the indices without a wildcard What if we automatically appended a wildcard as soon as they start typing? A user could remove it if they'd like but at least the initial experience would work as expected. Otherwise, we are specifying a single index which I believe is used much less than an index pattern with a wild card (but may be wrong). I'm not sure we want to advertise the optional ID so much. We prefer users use our unique ID. Can we add an advanced settings option here to hide it a bit? I believe this was in the original PR. I'm not sure "Optional Time filter field name" is the best header here. Especially given the fact that making a selection here is a blocker for me moving forward. Also, the option to "not use a time filter" is pretty hidden. I wouldn't think to click into this drop down to create an index without a time filter. This is an issue with the current index pattern creation view as well. What if we called the title "Select time field" where the dropdown provides all available options, with a checkbox that would disable this drop down and allow a user to create an index without a time field. @cjcenizal @snide I believe there is some history here for why this option is in the dropdown. Any additional info you can provide or suggestions for a better approach? There have been a number of designs from @alt74 around the index pattern management page for K7. On the design side of things, we seem to be in agreement that this information architecture is a good approach. There are still some kinks to work out but at a high level, the single step / side by side approach is preferred. What are your thoughts around the work effort to implement for K6? I'll set aside some time early next week to discuss in more detail. cc: @skearns64 |
@alexfrancoeur To provide some context, this PR currently doesn't function properly because of a technical issue. My suggestion is that we wait for @chrisronline to solve that issue, and then assess whether there are any design problems which need to block this PR from being merged. If this design is one which can logically evolve into where we're headed with K7, then I think we should merge it, and then open up discussions about design improvements. The longer this PR is outstanding, the more likely it is to be delayed by merge conflicts, the longer we have to wait on merging, etc. creating a vicious cycle. Unless this PR creates an objectively negative user experience, I think we should merge it and iterate -- progress over perfection. 😄 |
As part of this PR, we should make sure labels are properly associated with their inputs, per #12863. |
@cjcenizal I agree wholeheartedly with progress over perfection, there are a few items in my feedback though that I feel are blockers to the user experience. I agree, once the PR is in a working state we can revisit. Also, I'm not too concerned about the IA change from K6 ==> K7 and am happy to see this naturally evolve. I just wanted to bring it up as a discussion point. |
Handing this off to #13454 |
Replaces #12689.
Changes
Changes to the UX
Loading indices
Searching for matching indices
Submitting the new index pattern
To-do later
Feedback
From @skearns64
From @alexfrancoeur