jan_rix's picture

double postings

It doesn't happen often but sometimes people post a spot they have previously put up. When these are only minutes apart and adjacent on the carousel, it is obviously an error but when further apart, see http://www.ispot.org.uk/node/105974
posted half an hour after the identical one www.ispot.org.uk/node/105965 and IDd by the poster twice, this seems something that perhaps the software could iron out. Surely there ought to be a way that the software could recognise the photograph details and reject it if it has already been posted, certainly within the 30day time-frame maybe? This could eliminate the error postings at the same time if it is possible to set up.

Reply

Comments

miked's picture

Yes this is something I

Yes this is something I pointed out in the early days of ispot and its on the long list of things to do (eventually). Its actually something that can get worse over time once you have loads of observations and want to go back and enter some older ones but not sure if you have already added them.
Could indeed get the software to look at file name and file size or some other combination of details. But of course this is not foolproof as you may spot something and rush to put on ispot same day but a few months later you may have forgotten that you already put it on and have renamed the file and might want to crop it down to put on just the main subject so neither the filename nor filesize are the same as the original.

jan_rix's picture

Thanks Mike

I appreciate there is a huge amount of work to do on the website, which is a really great site and this is only a small matter. Maybe the software could do it by recognition of the photo primarily, rather than by other data? Its not a big problem really, and of course people do make errors especially over time, but when examples are half an hour apart? It seems like someone trying to boost their observations tally! There are also increasing numbers of double IDs going on now, identical IDs, either on their own spot or someone else's which has already had a correct ID placed sometime before (not within the normal timescale of coincidence of simultaneous posting) such as : http://www.ispot.org.uk/node/108051 and http://www.ispot.org.uk/node/107957

Dryopithecus's picture

Duplication may be intentional

I have an image of a bug attacking a ladybird pupa, so, when I can get it to upload, the plan is to post it twice, once for each of the two subjects. Do you intend to outlaw this procedure?

Thinking about it, this would be no real obstacle for me, as I could easily recrop the original or save it with a different compression ratio to give a second one. Some users may not know how to do this, however.

It may not be necessary or desirable for the software to delete what it sees as a duplicate posting. Presumably you have censors to check for abusive language or indecent photos & the like, so all that is needed is for a flag to be set to indicate possible duplicate postings. This flag could be set by the software, or users could set it for you by clicking on a button.

My suggestion is that the "Report content as inappropriate" button be adapted to encourage users to report duplicates as well as other inappropriate content.

Dry.

Refugee's picture

It happened to me!

I investigated it from my end.

What i have been doing is to use the "back" button on my browser between observations and the thumbnail page in order to avoid waiting for the gallery to download again each time. I do this for another cost related reason to do with ISP usage too. If i post an observation or comment and get distracted i can forget that i have saved it and see the un-saved copy in my browser and "think" i have not saved it. It is very easy to save it again leading to an un-wanted duplicate.

A reminder not to use the back button in your browser that appears for a couple of seconds after saving would get rid of many duplicates of this kind. There is nothing wrong with using it the rest of the time apart from on the home page where it always re-loads the whole page anyway.

If i realize that i have done it i will press the report button on the duplicate and add the reason i want it removed in the "reason" box.

Refugee

Martin Harvey's picture

double identifications

I agree with you that ideally identifications should not need to be repeated, and instead people should click on "I agree". However, in some cases the repeats arise because people studying the OU's "Neighbourhood Nature" course are required to add their own identification to a set of ten observations on iSpot. Ideally they should try to make an identification at the time of posting the observation, even if this ID is subsequently corrected by someone else, but this doesn't always happen.

For more see:
http://www.ispot.org.uk/help-etiquette

----
Entomologist and biological recorder

jan_rix's picture

call me sceptical but

I don't understand how borrowing someone else's identification can really help towards an OU course, except on paper. I have seen quite a few of these recently, and you are right, they are people on OU courses. I wonder if their tutors check whether they have made the ID themselves!

Tony Rebelo's picture

iSpot bug

iSpot should not allow a double posting of an ID (based only on the Scientific Name: Common Names do not count to an ID - esp. in southern Africa where we have 11 Official languages and competing groups with their own "Official" common names).

iSpot should catch a duplicate ID and inform the user that it is already posted and that they should agree with the previous ID or else post a comment should they wish to elaborate or discuss an issue about the ID.
A problem though is that iSpot sometimes links to the wrong dictionary and then a duplicate is needed to change it from one dictionary to anotherI (e.g. for an observation accidentally posted to Egypt - by leaving out the minus sign on the Latitude, then identified and linked to the CoL dictionary, then repatriated with a locality correction, and then not visible in southern Africa because the region uses the SANBI dictionary).
This could of course be fixed if programmers checked the dictionary if a the locality was changed and made appropriate corrections, but they do not seem to have realized that such an event might happen. Unfortunately it happens a lot!