> but surely that's worth it to save gigabytes of memory
I'm not sure _save_ is the right word here...using available memory has no particular cost to me over not using it. Of course, the _available_ modifier is relevant, so I take your point.
It's a huge advance for those of us who usually have 50-60 tabs open for weeks at a time, where hitting a really heavy image page would tip Firefox over the top and crash it.
I switched from Chrome to Firefox a couple of months ago for its ability to handle more tabs without slowing/crashing, and for its better performance with image-heavy pages. So, for me, it's not just "who cares?" it's a killer feature.
Why should one use bookmarks rather than many tabs? I have many concurrent projects going, and things I refer to regularly, so I just leave them grouped in my tree of tabs on the sidebar. Much more efficient than digging through nested bookmark menus, or trying to make sure the bookmarks are properly classified.
Basically, my tabs are bookmarks, except tabs are integrated into the web-browser better.
If they would make the tree-style tabs very aggressively swap to disk, or even discard state entirely and seamlessly become a bookmark that would reload when clicked (perhaps this behavior could work only for white-listed domains, like wikipedia.org), that would be perfect.
The main usability problem with bookmarks over tabs AFAIC is that the bookmark UI does not encourage pruning bookmarks, so most become stale over time.
Another idea I've been kicking around for a while is "tab decay". Have tabs start to decay and close themselves after not being viewed for a configurable amount of time. 1 month would probably do it for me, it would save me a lot of cleanup time.
Yes, that is true but there is no problem with just 60 tabs. I have 260 open tabs and my browser is not sluggish and the memory usage isn't too bad either, 3GB.
And then we have people complain that their "un-used" tabs are taking up too much memory, so browser makers cut down on the level of state stored in the tab and that brings us to the current point.
So store the state on disk if the user hasn't been to the page for a while. When I've I last visited a tab three days ago I don't mind a little disk delay when I get back to it, in exchange for not having disk churn everywhere when I'm trying to get work done.
So do users. Keeping several dozen tabs open indefinitely decreases the stability of every browser. Computers have resource limitations. 95% of websites have bookmarkable links; use them.
Not Firefox, which is why I use it instead of Chrome. I am not saying tabs are without problems but they fit way better into my usage patterns than bookmarks.
> I'm not sure _save_ is the right word here...using available memory has no particular cost to me over not using it.
The browser not using available memory means that other processes on the system have the memory available to use. For those of us who may have the browser running while something else is actively running on the computer, that's potentially pretty important.
You would be right if Firefox's memory usage was opportunistic, like the OS disk cache. Problem is, Firefox (or any browser for that matter) will tend to bloat in memory as it sees fit, which for any device with 4GB or less of RAM may mean swapping to disk, which is still the same killer as it always has been.
4GB seems to be the threshhold right now. Windows consumes maybe 0.5-1GB, budget 0.5-1GB for a handful of other programs, bloat Firefox up to 1-1.5GB, and suddenly you're swapping. (Naturally the OS does not wait right up until you have 4GB of RAM consumed to start swapping, it seems to start around 3GB)
I'm not sure _save_ is the right word here...using available memory has no particular cost to me over not using it. Of course, the _available_ modifier is relevant, so I take your point.