Currently records by default are sorted by most recently added. I would propose the default order be switched to most recently updated.
This way if a comment, an ID that does not result in RG status being attained etc are entered, the record moves up in order where it is more likely to be reviewed.
This would be especially helpful for any efforts to clear the backlog of approaching 210,000 pages of needs is records.
Often someone will add a comment or ask a question, that can lead to another line of investigation that leads to a more refined ID. Having those brought back in front of the active IDers is a great idea!
although⌠would they still be marked as reviewed? I can see problems with messing with that flag⌠currently we can tag others that might find our comments useful, and we have the alert system for when we are following an obs. If we start clearing the reviewed flag of other identifiers, we would be massively increasing the workload!
No, reviewed observations would still be marked as reviewed. The main idea is to bring to the front of the list things that were observed a while ago, but were recently identified as being in the taxon of interest.
I changed the title of the topic to be more explicit, so please correct it if I got the gist wrong. While we could do this, there are a lot of different reasons an observationâs updated_at date might change, not all of which imply a need for more IDs. For example, adding an annotation changes updated_at, so does just editing the obs to change the description, as well as DQA votes, adding to a traditional projects, and any number of background tasks.
How about we add it as an option and not the default and see if itâs useful?
I should also note that âclearing the backlogâ of Needs ID observations isnât exactly a reasonable goal in my personal opinion. Many observations canât be IDâd and will probably sit in Needs ID forever, or until the state of our knowledge changes (e.g. we discover that red palps on jumping spiders in County X always means species Y). I think we should all be doing our bit, and I think reviewing everything in within particular search parameters is laudable, but actually having nothing in Needs ID is a bit lofty.
Also, why not just filter by oldest first? I started doing that and while thereâs tons of tricky stuff I donât know thereâs also lots of neat tricky stuff itâs fun to see. Combining that with my âcurated plant observersâ project, Iâm paging through a bunch of weird neat plants @kueda added in central California and a bunch of sedges from @bouteloua âs old observations. No I wonât get through them all but it helps dig up some neat gems. And while I do it rarely one can also mark observations with multiple higher level IDs as not needing further ID. The random sort also is fun. So I personally probably wouldnât use this newly proposed filter.
Well as an example there are 3,000+ pages of plants in Canada awaiting review. Iâm far from a plant guy, but say I can get some of these into a family or genus. If I can do it for something that is on page 500 (or say I just do my local county, and those observations come in at page 500), by having the sort on recently updated, when the plant experts come through, it does not rely on them getting to page 500.
Right now it kind of feels like if
I canât identify it to species
or there is a risk the submitter is inactive
There is no value in me even trying to help on older stuff due to the risk it stays buried hundreds of pages deep and my attempt to help will remain unseen.
@kueda - yes I know having nothing in the ID pool is an unrealistic goal, reducing may have been a better verb of choice.
I donât feel this way. I add IDs when I think I can express what the right ID is. If I canât ID it to species, maybe I can improve it by adding a finer ID. If that doesnât shift the taxon the observation is associated with, then maybe someone will follow up later and agree with me (or disagree). If the observer is inactive, so be it. More IDs will still help, even if theyâre inactive and theyâve opted out of the CID, b/c a majority of disagreeing IDs will boot it from Needs ID. I think it also just helps for people to see an ID is contested.
Anyway, do you want sort by updated_at as an option? Anyone else?
@kueda I would be interested in seeing what it can do, assuming it is an easy add. I kind of misinterpreted it a little with my first comment above, thinking it would be like a kind of automatic alert effect, but for that to happen it would need to be marked as unreviewed for it to be back into the Needs ID pool as far as those who have already seen it are concerned.
I see the observation as being the gold, not the ID itself. The ID just âadds valueâ to the observation, and it is not worthless without it, and an ID could be added at any stage in the future! We are literally dealing with âdynamic IDsâ here, and just as they can be added years down the track, so too they might change years down the track! When a new expert jumps in, say lepidoptera, then having things IDd to that higher taxon will likely put it into their review stream, regardless of how long ago it was!
âDate addedâ is still the default filter, so as long as youâre happy with âdate updatedâ not being the default, sounds like closing would be good.