When trying to flip to the next observation, only the name changes

Bug Reports
19

i think the symptoms of the problem are more likely to surface – at least temporarily – if you have a slow connection, but the original description of how to trigger the issue must be wrong.

i don’t think you can reproduce the issue by using the green arrows on the screen. i’m fairly certain you have to use the browser’s forward and back buttons.

i tried to replicate the issue using only the green arrows and i was not successful. here’s what i see:

  1. the green previous / next buttons become available only after the screen has issued requests to determine the next 6 and previous 6 observations and received valid responses. (if valid response is not received from both requests, the buttons never become available.)
  2. if you mouseover the previous or next buttons, you’ll get little pop-ups telling you what the taxon of the previous and next observations. (this is because the screen already knows what these previous and next observations are from #1 above.)
  3. if you go offline / disable your network and then click the next or previous button, you’ll get the basic outline of the next or previous observation with all the text but missing any photos or maps. (this is because the screen already has all the basic details about the observation from #1.)
  4. if you’re online and click the previous or next button, the screen doesn’t even make a request after your click to get details for the previous or next observation (because it already has details from #1)

this tells me that there can’t be some sort of race condition when using the green arrows, and there can’t be some sort of failure to get infromation for the previous / next observations using the green arrows (because if the arrows are available, then those observation details have already been retrieved). i think that means it’s impossible to recreate the symptoms of the problem by using the green arrows.

however, if i go back and forth between two observations using by browser’s forward and back buttons (while i have my browser throttled to 3G speed), i can see the signs of a race condition happening there, similar to what’s described in the original post. if i wait long enough, it seems like the page will usually load fine in the end. it’s possible that maybe if some of the requests fail or complete in the “wrong” order, you’ll end up in the transitional state where information from both observations get mishmashed together, but i didn’t try to get to that point.

i think the main point that i wanted to clarify first was that the triggering of the issue must happen with the browser’s back / forward buttons and not the observation’s green previous / next buttons. is that what others see?