dside

@QuadratureSurfer best I can tell, per every instance of the page open. So roughly per user.

The requests are made straight from viewers' browsers to ECI.

@QuadratureSurfer @skg looking at the code of the progress tracker, the tracker website makes direct requests to ECI every 3 seconds without even waiting for the previous requests to finish{0}.

The large number of people keeping this tracker open could very well be the cause of the overload:
https://github.com/Alex1337F/StopKillingGamesTracker/issues/25

Closing the progress tracker and maybe reducing its exposure a wee bit until it's fixed sounds like a good idea :blobcateyes:

Maybe the team behind ECI should have expected this kind of load, sure, but it's what we have today and it'd be a shame to lose the initiative to this :blobcatfacepalm:

edit: oops, looks like Lemmy misinterpreted my notation for "footnote links" out of square brackets and rendered this post in a broken way, let's see if curlies help

edit²: there is a PR that intends to stop new update requests from firing when an earlier request is unfinished: https://github.com/Alex1337F/StopKillingGamesTracker/pull/26 — which should help *somewhat*

{0}: https://github.com/Alex1337F/StopKillingGamesTracker/blob/master/main.js#L972-L973

@besmtt the only right answer to this question as *meant*, not as *written*. :blobcatfingerguns:

@NONE_dc

@mittorn возможно, обычные задержки федерации, посты по федерации безотносительно групп доходят не мгновенно, а очереди с ними могут обрабатываться в разных темпах, из-за чего порядок появления может быть не хронологическим. Что запутывает, конечно :blobcatsad:

@linuxmemes

@mittorn я на @linuxmemes не подписан, но на странице группы на MML1 как минимум твой пост имеется.

@mittorn щёлкнуть в меншн, попасть в профиль и нажать кнопочку "Подписаться"?

@linuxmemes