Meta-nonathlon 2020-2021

[JonT]

Ok. I’ll take a look later this week. Thanks for pointing this out

[JonT]

This is a really odd one, and unfortunately is being made difficult to fix because of the "observer effect".

I can see in our Nonathlon Analyser database that your score had not been updated, which is why your Metanonathlon score was incorrect - it was using the old score from the database, not the new one from the Nonathlon site itself.

The database should be updated if your scores have changed whenever anyone looks at the analyser - which clearly hadn't happened.

But.....as soon as I went into the analyser to check it spotted that the scores were wrong and updated them i.e. my observation had an effect which now means I can't see the error any more

So the question is, why didn't this happen when you entered the analyser?? Hmmmmm. And is it a coincidence that it was the 500m on both occasions? I suspect the answer is yes to the second question.

[Grobi]

Thanks for fixing this Jon, although we still do not completely know what the problem was/is. Maybe the analyser only accepts your command to update

The 500m were pure coincidence. I updated my hour yesterday with the same (non) effect but didn‘t want to bother you again.

[JonT]

Thanks for fixing this Jon, although we still do not completely know what the problem was/is. Maybe the analyser only accepts your command to update

The 500m were pure coincidence. I updated my hour yesterday with the same (non) effect but didn‘t want to bother you again.

Please do bother me. That kind of information will stop me going down dead ends. I'm having another head scratch about it right now. It's bugging me!!!

[JonT]

It has been a while since I put an update here rather than leaving people to look on the site.

Paul has shot up from mid table to 2nd

Peter and Steve have also made some ground.




Sent from my iPhone using Tapatalk

[JonT]

Thanks for fixing this Jon, although we still do not completely know what the problem was/is. Maybe the analyser only accepts your command to update

The 500m were pure coincidence. I updated my hour yesterday with the same (non) effect but didn‘t want to bother you again.

Peter - when you access the Nonathlon Analyser and Metanonathlon, how are you doing it? Are you following the links in the utilities menu for example? I'm just wondering whether there is a way to access the page that is causing it not to update i.e. I am still scratching my head

[Grobi]

Jon, I use the link in the utilities menu for the metanonathlon. I access the nonathlon analyser through link on the nonathlon page.

[JonT]

Jon, I use the link in the utilities menu for the metanonathlon. I access the nonathlon analyser through link on the nonathlon page.

That doesn't explain it then

Looking at the logs, your scores were updated recently on the following dates:

2020-11-06 21:12:13 - Updated scores for Grobi
2020-11-09 22:00:15 - Updated scores for Grobi
2020-11-12 20:15:07 - Updated scores for Grobi
2020-11-18 14:35:57 - Updated scores for Grobi (this is the one triggered by me looking at the analyser yesterday afternoon)

Do you know the dates when you entered the updated scores into Nonathlon? They should pretty much match the dates above. The times will probably be off because the update isn't triggered until someone views the analyser itself. (The times are also GMT).

[Grobi]

The update dates you found are correct:

Nov 6th: 10k
Nov 9th: 60 mins
Nov 12th: 6k

I‘m not sure about the exact time of day but it was definetely in the evening, so they should be correct.

On a positive note, I just updated my nonathlon score for 30 mins. It‘s now 876 (was 870) and the metanonathlon updated from 100 (870/866) to 101 (876/866).

I had thought that maybe my german ip address was the problem, but we can rule that out now.

[JonT]

I sometimes find the colour coding of the Meta-nonathlon a bit confusing. It gets most confusing around 100 points where it is possible to be yellow (exactly 100), orange (above 97.5 and below 100) or light green (above 100 and below 102.5). To help a little I have changed the display, if you hover your mouse over a cell it will show the actual score to two decimal places instead of the rounded score which is displayed in the table.

P.S. I still haven't found the intermittent error where scores are not updated. Anyone who writes code will know that when you see a comment in the code like "No need to do X because Y will have taken care of this" your instincts are that the developer knew this was possibly not true, otherwise why mention it in the first place!

I just found such a comment in my code, basically saying that there was no need to update the score database because the Nonathlon Analyser would take care of that. I now wonder if that is actually true, or whether in the back of my mind I wasn't really convinced. I'll keep digging.

[Iain]

Many thanks for all your hard work Jon. My 1k updated fine yesterday. I knew that I was unlikely to catch you, but blowing my dreams away the very next day was almost cruel

Not a big deal, but could you add this feature to the total as well, would be good to see how close we are to those around and whole percentages don't tell the full story.

Thanks - Iain

[JonT]

Not a big deal, but could you add this feature to the total as well, would be good to see how close we are to those around and whole percentages don't tell the full story.

This wasn't quite as trivial as I had expected, but should be working now.

I don't think the result is what you were hoping for

[Iain]

Not a big deal, but could you add this feature to the total as well, would be good to see how close we are to those around and whole percentages don't tell the full story.

This wasn't quite as trivial as I had expected, but should be working now.

I don't think the result is what you were hoping for

You were away and clear with your excellent PB anyway , so no expectations, just thought it would be a useful feature. Many thanks for adding this.

[JonT]

P.S. I still haven't found the intermittent error where scores are not updated. Anyone who writes code will know that when you see a comment in the code like "No need to do X because Y will have taken care of this" your instincts are that the developer knew this was possibly not true, otherwise why mention it in the first place!

I just found such a comment in my code, basically saying that there was no need to update the score database because the Nonathlon Analyser would take care of that. I now wonder if that is actually true, or whether in the back of my mind I wasn't really convinced. I'll keep digging.

I have made a change to the Meta-nonathlon code so that it checks the Nonathlon database is up to date before displaying the Meta-nonathlon table. This may be the issue that was causing the intermittent problem of scores looking out of date in the Meta-nonathlon while they were up to date in the Nonathlon Analyser. As ever, please let me know if you are still seeing this issue.