Issue uploading log

I can’t manage to upload my latest log - yesterday’s activation of SP/SP-001.

I press Submit after veryfying the log (no errors) and it goes in a infinte loop (stuck on screen below).

The log still gets somehow uploaded (I see it in my stats) but without any points awarded. I’ve tried a couple of times after deleting the log. Am I missing something obvious? Note that I didn’t activate this summit yet this year.

1 Like

I’ll take a look through the logs later tonight. If you could forward me the CSV/ADIF to anryan at Gmail I can more easily debug what is going on

1 Like

Sent. Thanks a lot Andrew.

1 Like

I have the same issue that SP6SUD has reported. My log is from Sep 3, OE/OO-033.
73 Heinz

2 Likes

I’ve checked quickly and it works fine for me, which is of course an unhelpful diagnosis! I will keep investigating but I am in a meeting for the next few hours.

It looks to be failing in the piece where it updates your activator score so it may be something specific to your accounts that isn’t specific to mine.

1 Like

No worries Andrew, it’s not urgent obviously :slight_smile:

One thing that may help understand the problem is why this activation show 0 points (I made 8 QSOs and didn’t activate this summit this year):

1 Like

And your activation show 0 points as well. Glad I am not alone :sweat_smile:

2 Likes

Also for me an issue to upload chaser qso. Same as previous days, generally solved in half a day. I’ll wait…

1 Like

In case this helps… tried twice with iOS and it wasn`t uploading the ADIF file. Then a try with Firefox on Windows and it worked without a problem…

73
Ingo

1 Like

My suspicion was correct! However, it is kind of obscure. All three of you (SP6SUD/SP9MA/OE5EEP) are now on the new DB API, which was failing on the fact that at some point this year, you’ve activated a summit with a winter bonus period, and in testing that summit for whether to include in the activator score, it was producing an invalid date. A quick check of the older version of the API revealed a mistyped initial value for one of the variables, which hasn’t impacted 95% of people using the new API, but hit you folks.

This is now fixed. To get credit for the activations, you can go to the upload, edit it, make no changes, just hit verify and submit and this will rescore your activations and trigger a badge update.

You, on the other hand, do not appear to be directed to the new API, so this is the common or garden instability we have seen on the old API. During this period of instability on the old API, the new API ticked along fine, so this is a good sign.

This is a different issue again on the frontend, caused by me fixing some undefined behaviour and causing another issue, which should be fixed now. Basically, the way each browser handles the numbers that are pulled from an ADIF file or a CSV file is different: Firefox interprets as a string, Chrome as a number, Safari as whatever it chooses (probably something ridiculous). So, to fix that to work with the new API (which cares a lot more about types being passed to it), I forced everything to be a string type, but in doing that, encountered another problem with certain ADIF files that required a two-line change.

4 Likes

Slightly different problem here: The upload gets stuck right after pressing “import” on Windows 10. I tried Firefox and Chrome, same results for the last 24 hours. Same result with FLE import.

1 Like

Same deal - send me the CSV and FLE files to anryan at gmail and I’ll see what I can do. This will be a frontend issue, and do try a refresh to see if the recent fixes were actually picked up by your browser.

1 Like

QSL - all good now after following your suggestion. Thank you for your service Andrew :pray:

1 Like

Yes, corrected it as you said and everything is fine now!
Thanks for your support!

Was there a reason why this bug showed up for the three of us on the same day?

73 Heinz

1 Like

Make that 4 people…

All working for me now too.

Thanks to all the volunteers who keep this all going.

73
Gerald

2 Likes

Yes, 2 things were needed.

  1. You need to have the kind of activations in your log to trigger the bug.
  2. You needed to be switched to use the new DB/API so you were exposed to the bug.

We have only just started switching people to the new DB/API. It’s being done in chunks so we can see the effect of load and also expose the code to the hugely varied data we have have collected in the DB.

People on the old DB/API have a set of intermittent bugs that have driven us mad but mainly annoyed Andrew. Of course there will be a different set of issues now :wink:

5 Likes

Happy to help finding them… :rofl:

2 Likes

Thank you very much!

73
Ingo

1 Like