The OpenScience Laboratory The Open University

iSpot Forum

BUGS, ERRORS and FAILURES

It looks like the links are all the secure (https) links but (many of) the sites in question may not have configured themselves appropriately for this.

https://www.mammal.org.uk/ produces an error as you describe. http://www.mammal.org.uk/ does not.

https://www.mammal.org.uk/ is what the link is to, http://www.mammal.org.uk/ is what many browsers will default to viewing if you types just www.mammal.org.uk into the address bar.

Supporting an https access and linking to the https version when connecting from an https site (like iSpot) seems like good practice to me but a depressingly large amount of websites do not support it. Here’s an old article about this https://www.wired.com/2016/03/https-adoption-google-report/

1 Like

Oh, and to think I used to believe I had a good working knowledge of IT! I presume there is some real advantage in iSpot linking to the secure versions, but unless the link actually works, it’s confusing. Maybe, in the fullness of time, the other sites will “catch up”, and the links will work.
Thanks for the explanation, anyway.

1 Like

Thanks…not many Likes being given, now we are so Jaded

@miked… c’mon ADMIN Wake up to the difficulties of PROJECTS - please?

The issue with projects, carousel and some other items on the site that don’t work seem to be down to a wrong bit of configuration that happened when trying to correct another issue a couple of weeks ago. There are a load of items that we know about and which might be fixable by other programmers but the key programmers that upload fixes and deal with configuration are working on a different project (not ispot) that seems to have a tight deadline and so can’t spend the time sorting out ispot. This is extremely frustrating for me as it is for all ispot users. We are trying to put as much pressure on to get some time for them to work on ispot just to sort out these config issues and upload the fixes but so far other project is taking priority.

1 Like

OK, thanks - accepted (by me anyway) It’s the long silences that are so demoralising…

yes - thanks Mike. It does help to know whats going on and the reasons why some things are taking longer to fix. Feel a bit more positive now.

but so far other project is taking priority.
Sounds interesting

Not that interesting as the other project has nothing to do with natural history, it is either something to do with psychology or the police I am not sure which as have not seen it, they do a wide range of projects.

Has anyone seen the wrong likely ID bug recently? According to the programmers this should have been fixed a few weeks ago but they did not realise at the time it had been fixed as it was a knock on effect of a different bug fix.

Are you talking about when a mere beginner’s ID is higher ranked than mine, as the expert? This is still there. And very frustrating as it means there are numerous little issues still lurking in the programming.

That is exactly what I mean “Are you talking about when a mere beginner’s ID is higher ranked than me, as the expert”
New cases like this should have stopped happening at least a couple of weeks ago, probably longer, I can’t remember exactly when the particular fix was applied.

Could you give me any examples of (a)new inistances of it still happening and (b) old cases where it still exists. I have asked if it is possible to run a database query to find all the old cases (so they can be fixed) but not sure when or if that will happen.
If it is still happening then of course further investigation and a new fix will be needed.

Yes of course. But it is difficult to find when another agreed to the new correct ID. Also, please correct me, if you correct or improve another expert’s ID, should the likely ID not also change? This is often not happening, meaning the two expert ID’s are not the same weight.

Unfortunately again we need to see examples. We spent ages working through a set of examples (getting all the actual points etc) that looked rather odd and it turned out that the system was working correctly in each instance. I did not believe this would be the case until we went through them in detail.

I have had a quick glance through your identifications and found this one:


which looks a bit dodgy. The ID date is around the time the change was made that should have fixed the bug so it may have been before the fix was implemented. I will use this one as an example for the programmers so don’t make any changes (and hopefully no one else will either altough it is possible to check points total in database at set times)

I have a case, https://www.ispotnature.org/communities/southern-africa/view/observation/666303/tiny-blue-crabs where (the crab being unlike the northern hemisphere P minutus as far as I knew) I added a new ID weeks ago, switching the experts ID, but finding P. minutus was found in S afr as well, concluded the local experts knew best, so had to agree to the 1st ID. So I’ve just now unagreed the 1st ID switching likely back to my less expert ID, so the bug does still exist!
And I can do the same with https://www.ispotnature.org/communities/southern-africa/view/observation/668805/southeaster-casualty-1 if you like.

Is it possible in the above, and other cases, that the area of ‘knowledgeable’ sways the result?. These are both marine inverts and Chris is ‘Knowledgeable’ in that field. I imagine that Tony is not.
The test, of score, could be made by Tony offering an ID inside his Inverts specialism - the result might be interesting and might suggest iSpot is intelligent - rather, dare I say, the Programmers!

One think is that I think iSpot is static and updated on changes. So once a ‘wrong’ ID has been locked on an entry by the bug I’m not sure it would be updated until a recalculation is forced by someone agreeing to an ID. Certainly I have in the past forced changed by agreeing and the withdrawing my agreement on an entry.

If I am right (and happy to be told I’m not) then where there has been errors fixing the bug will not correct the old ones (so they will still exist with the expert’s ID overridden) until there is a change. But if the developers are right no new ones should occur.

Well the whole point of my cases shown above , https://www.ispotnature.org/communities/southern-africa/view/observation/666303/tiny-blue-crabs & the other, is I can switch likely ID any time I want just by agreeing or unagreeing Tony’s ID; & yes tony does have a ‘knowledgable’ though it’s pretty hard to see.

I see the carousel bug is still around. I ID’d a moth ob https://www.ispotnature.org/communities/southern-africa/view/observation/751927/moth then decided to check some of the ’ other observations of… that had been generated, clicked on the first one (which I recognised as possibly being one of my own) and oops what a lovely yellow Iris popped up https://www.ispotnature.org/communities/southern-africa/view/observation/747863/moraea-longifolia This ongoing bug makes the whole sight pretty useless

Latest inconvenience - when making an observation you can’t access the Identification page until all the image had loaded. (Previously, I understood, you got held at the Confirmation page.)