We're suffocating on inventory space - how much of a priority is it to fix it?

[BreathlessTao]

They mentioned for example that they could backup stuff less often, but that would mean higher risk of losing items in case of a server crash.

They need to back up stuff less often. That ~15s is totally insane: on one hand it's flooding the network with traffic (at least, I'd guess, half of which is completely unnecessary) continuously; on the other, it's a strain on the servers themselves - no idea what they're using, but if it's SSDs it's certainly a quick death for them, if it's HDDs I wonder if they can even finish one save before starting the next. There are simply too many characters online at any given time, even if only half of their inventories is full, it's way too much data. They even let AFKers just idle away and do nothing for 6-8-10-12 hours, dozens of them at any given time in any city, and their inventories are being backed up too.

And for what? That 90k disconnect error is as frequent as ever. Network strain and data sizes and server limitations are mentioned in pretty much every live letter, in pretty much every issue we get a negative reply to (which is quite a lot). While at the same in the 2+ years since ARR launched - how many servers crashes did we have? How many times did we need a rollback? As far as I remember: zero. That backup timer need to be increased, at least to 1 minute - that would be much more rational on both sides of the argument. (Personally, I'd say once every 5 minutes, with logout and retainer exchanges actively triggering a backup.) Heck, have the acquisition of blue/purple items (even scrip-exchanged items, for all I care) actively trigger a backup. And for sanity's sake, put the AFK-kick feature back in and don't ever remove it - if you're away for more than an hour, if you're "stuck" in "crafting" or in a "conversation/cutscene" for more than an hour without doing anything, you've no place to be online in this game, not with the systems (backup and hardware) that it currently has. It's a very remarkable system, in a way (not in a good way), that's so damn spartan and strict on one side of an issue, and so utterly lax and free on the other side of that same issue...

[Nekodar]

They need to back up stuff less often.

The backup mentioned is done server side. There is no synchronisation done between client and server that often afaik, so the backup doesn't mean network flooding.
Inventory related network traffic is generated when you change anything in your inventory, and when you go into instances.

The 90K disconnect errors are generally not on SE's side of the equation. ISP's and backbone providers are the main cause of those.