Project

General

Profile

Task #6348

Story #6330: Address feedback on MetacatUI features and usability

Address feedback from Carl Boettiger

Added by Lauren Walker over 8 years ago. Updated over 8 years ago.

Status:
Resolved
Priority:
Normal
Assignee:
Target version:
Start date:
01/03/2014
Due date:
% Done:

0%

Estimated time:

Description

Hi KNB team,

writing my impressions after a quick test drive of the new KNB interface.

Testing on Google Chrome 31.0.1650.63, Ubuntu 13.10

1) Impressions, issues:

Really slick, awesome interface. Great work. I love the auto-complete/matched list boxes on the filter. A few minor comments:

- This is probably me being stupid, but I expected to be able to find this file: https://cn.dataone.org/cn/v1/object/knb-lter-hfr.205.4 on the KNB. Searching by author, identifier, etc, doesn't seem to give me any results.

- After entering a search query and hitting enter, the search box is cleared before the results appear. It's strange not seeing my search terms still in the box, and for a moment makes me think my search didn't happen. With map off this is fast enough not to be so confusing.

- Mapping feature is very nice, fast, responsive. Took me a while to find the hide-map option (long enough to originally request it as a feature -- whoops!)

- Filter box shows horizontal and vertical scroll-bars even though it doesn't scroll

- Filter box appears partly covered by the "hide map" and "share data", particularly when map is open (at least on my smallish laptop screen). I would actually remove/relocate these buttons entirely.

- I would have put the "filter" option attached to the search results box, or as an "advanced search" option? Took me a while to realize what/where this option was.

- the "share data" slide out is cool, but become annoying as it pops up unintentionally when I have the mouse in the filter-box area, and it isn't where I'd click to share my data (The "share" button on the top menu being 100 times more obvious...)

- It is unclear how the map is handling bounding boxes or ranges of many sites, since it appears that each dataset is represented as a single pin in the map.

- Layout of metadata summaries looks very sharp. However, blank boxes often appear -- it seems more logical to simply omit the boxes when they have no value entered. (e.g. attribute information here: https://knb.ecoinformatics.org/m/#view/wolkovich.33.3)

- I was a bit surprised to find coverage metadata appear only after the data table metadata, especially when there are many data tables: e.g. https://knb.ecoinformatics.org/m/#view/doi:10.5063/AA/wolkovich.3.16

- In general I like whitespace, but some of the coverage metadata appears too spaced out to read easily. In particular, taxonomic coverage would be much easier to read in a tighter table than in boxes of "Classification" with the words "Rank Name" and "Rank Value" appearing again and again (I realize that EML's taxonomicCoverage isn't always just rank name-value pairs, but when it is it would be preferable to display them in a more natural layout).

- The taxon search can be misleading: if the EML only lists genus/species name, say, but I search for a higher taxonomy, I don't get results that should otherwise be matches. (e.g. this file https://knb.ecoinformatics.org/m/#view/esa.46.5 does not come up in a search for taxon: Sarraceniaceae.

- The URL doesn't show the data set I am viewing, or the search/filter terms I have selected. (If I visit the file by clicking on the DOI, I do get an informative URL, just not when browsing to it via the interface).

2) Suggestions for future features:

I wonder if you might consider browsing by other Coverage axes -- e.g. how about showing the data along a global timeline, or distributed across a tree of life?

Thanks again for making such a great interface. Apologies for the earlier email without the review content, and let me know if I should be sending this information to something other than metacat-dev.

Carl

History

#1 Updated by Lauren Walker over 8 years ago

  • translation missing: en.field_remaining_hours set to 0.0
  • Status changed from New to Resolved

1) Impressions, issues:

Really slick, awesome interface. Great work. I love the auto-complete/matched list boxes on the filter. A few minor comments:

- This is probably me being stupid, but I expected to be able to find this file: https://cn.dataone.org/cn/v1/object/knb-lter-hfr.205.4 on the KNB. Searching by author, identifier, etc, doesn't seem to give me any results.

This particular data package's replication policy is set to false

- After entering a search query and hitting enter, the search box is cleared before the results appear. It's strange not seeing my search terms still in the box, and for a moment makes me think my search didn't happen. With map off this is fast enough not to be so confusing.

We can keep the text in the text inputs until the 'current filter' node has been created, but I am hesitant to keep the text in the input longer than that because users could also interpret that behavior as their text not being submitted. I am thinking we can instead have an animation of the currently-applied filter nodes moving down from the text inputs as a clear movement outside of the input and into the list of current filters. Tickets 6350 and 6351

- Mapping feature is very nice, fast, responsive. Took me a while to find the hide-map option (long enough to originally request it as a feature -- whoops!)

We can move this button if it seems like there is a consensus that it is hard to find

- Filter box shows horizontal and vertical scroll-bars even though it doesn't scroll

We can easily hide the horiziontal scrollbar but the vertical one will always be there because at certain page heights the scrollbar will be necessary. The filter sidebar has a max-height so at some point the scrollbar will come in handy. Ticket 6352

- Filter box appears partly covered by the "hide map" and "share data", particularly when map is open (at least on my smallish laptop screen). I would actually remove/relocate these buttons entirely.

I will move the hide map and share data buttons up the page a bit to give the Filters sidebar some more room. When we add more filters in the future the design of the sidebar may be changed a lot, so we can save a total redesign for down the road. Ticket 6353

- I would have put the "filter" option attached to the search results box, or as an "advanced search" option? Took me a while to realize what/where this option was.

Noted. I think we may move in that direction if the sidebar gets too crowded in the future or if this feedback is common

- the "share data" slide out is cool, but become annoying as it pops up unintentionally when I have the mouse in the filter-box area, and it isn't where I'd click to share my data (The "share" button on the top menu being 100 times more obvious...)

This is annoying, and will be fixed with ticket 6353

- It is unclear how the map is handling bounding boxes or ranges of many sites, since it appears that each dataset is represented as a single pin in the map.

I can add a question mark icon to the 'Using map boundaries' currently-applied filter that will popup a more detailed explanation. Or perhaps a better term than 'Using map boundaries' can be used. Ticket 6354

- Layout of metadata summaries looks very sharp. However, blank boxes often appear -- it seems more logical to simply omit the boxes when they have no value entered. (e.g. attribute information here: https://knb.ecoinformatics.org/m/#view/wolkovich.33.3)

We can edit the xsl files in Metacat to skip over attributes that don't exist, but for now we will leave it. I added a ticket to the Metacat backlogs for this if we have time to get to it. Ticket 6355

- I was a bit surprised to find coverage metadata appear only after the data table metadata, especially when there are many data tables: e.g. https://knb.ecoinformatics.org/m/#view/doi:10.5063/AA/wolkovich.3.16

We can change this in the Metacat xsl's if it looks like it is a common suggestion

- In general I like whitespace, but some of the coverage metadata appears too spaced out to read easily. In particular, taxonomic coverage would be much easier to read in a tighter table than in boxes of "Classification" with the words "Rank Name" and "Rank Value" appearing again and again (I realize that EML's taxonomicCoverage isn't always just rank name-value pairs, but when it is it would be preferable to display them in a more natural layout).

This is something I have noticed too, especially when the taxonomc info is particularly long, it runs off the page and requires horizontal scrolling. I created a ticket for 1.5.0 - Ticket 6356

- The taxon search can be misleading: if the EML only lists genus/species name, say, but I search for a higher taxonomy, I don't get results that should otherwise be matches. (e.g. this file https://knb.ecoinformatics.org/m/#view/esa.46.5 does not come up in a search for taxon: Sarraceniaceae.

This will be resolved when we do more work on semantic search

- The URL doesn't show the data set I am viewing, or the search/filter terms I have selected. (If I visit the file by clicking on the DOI, I do get an informative URL, just not when browsing to it via the interface).

I am assuming the idea here is to be able to copy the search URL and view the same search again later? We have thought about implementing that feature in future metacatui releases

2) Suggestions for future features:

I wonder if you might consider browsing by other Coverage axes -- e.g. how about showing the data along a global timeline, or distributed across a tree of life?

Also available in: Atom PDF