Locations Plugin

Heh - yes - I didn’t want to go that route specifically for that reason. :slight_smile:

2 Likes

I’ve changed mine…

It produces this effect:

3 Likes

The unintended consequence is found in the Users List:

3 Likes

Sorry for the delay.

There’s no API for this afaics.

It’s just taking the value from the title:

https://github.com/paviliondev/discourse-locations/blob/c947297e8177f751b0f92dc15c52baecd6d4d75f/assets/javascripts/discourse/widgets/map.js.es6#L93

2 Likes

Has anyone come across this before? that the map just occupies only a portion of the map space? It always displayed just fine, until we updated the plugin.

1 Like

Does this resolve if you remove the Layouts plugin?

2 Likes

Oddly, the map doesn’t show up (all blank) when I remove the layouts plugin. I was originally thinking it might be a CSS issue but the issue persists after I removed all CSS.

Which is real odd, because your map at Pavilion looks just fine.

1 Like

Cannot repro. Check for a Theme clash?

Latest update to core seems to have killed this stone dead:

image

Clicking the user map, or the map icon on a user card causes a self inflicted dos :pensive:

It then just keeps going, killing the browser stone dead within a handful of seconds:

dead2

Same on the user map page:

I’ve had to disable the plugin completely this morning.

1 Like

Related perhaps?

1 Like

Sorry, not going to get much time to look at this right now, moving house and outside the support period double whammy.

I would first check this is not something else which is impacting the server, it may not be Locations at all …

I cannot presently see anything going wrong with Locations on current build, there’s nothing in logs.

Please take a look at what might be stealing performance.

1 Like

I’d say it’s probably the 4,000+ network calls being made to map.json every time someone tries to view a users location :smiley:

No worries, I’ve disabled the plugin for now.

2 Likes

Working fine on my instance. Just updated. Noticed a load of uncached resources at first. Fine now. To do with redis upgrade?

1 Like

I genuinely don’t know :man_shrugging:

Would that cause the thousands and thousands of network calls though?

1 Like

Every single avatar for all the users in scope? (for example). Every piece of map required.

1 Like

No, sorry, I meant, would it cause the thousands of network calls per minute to map.json that I was seeing?

1 Like

A 429 that’s being ignored, perhaps? If an asset is needed again, but denied, it may still go after it.

Yeah, not sure about the 200’s … what provider are you using?

1 Like

When I clicked on a user card, then on the map icon for that single user, it instantly started firing hundreds and hundreds of network calls to map.json

Over 1,000 calls to map.json were sent in under 60 seconds, just from my browser alone.

To the point where my desktop PC browser locked up completely and when I opened my discourse on my phone I instantly got the “slow down, too many requests from this IP” message.

1 Like

Probably more, but my browser died at that point :laughing:

1 Like

I tried again.

Same thing - kills itself instantly:

And:

etc.

On the off chance it was browser related, I tried Chrome instead of firefox.

Chrome seems to handle the issue slightly better and it got to over 10,000 network requests to map.json in under sixty seconds, at which point my entire PC started locking up - not just the browser:

Don’t worry about it @merefield I’ve turned it off again :slight_smile:

1 Like