01-06-2023 04:54 PM
adding more memory is not a solution because I know from experience that adding memory in comparable situations only "postpones" the problem (it might be in 6-10 MSN tabs for example in this case).
I also observed many years ago (maybe 20 years now - 32 bit world) that the way an application accessed memory could have a dramatic effect on swapping. one application I worked with on a P3 700 took 2-3 hours to run a test with 2GB of memory, was well on the way to 24 hours or more with only 1GB (and the swap disk maxed out) but with a simple "tweak" to the code (a performance "optimization" was actually removed) it took 3-4 hours with only 256MB of memory. the test was designed to max out the process memory (a 1GB process limit on that system).
I have been a computer guy since the 1970s and remember tweaking code and memory usage to make systems work within limits of 64K systems (eproms of 2K or 4K for example) and have worked on applications which made a Quad Core Xeon 16GB crawl due to swapping issues. I do not consider myself a "power user" (if I did I would probably have an I7 32GB as a low end system to play with)
I am not sure this qualifies as a bug (I would call it a bug though). I think it qualifies as a performance issue but I would think that I am not the only one to experience an issue like this opening only a few tabs so I would suspect that this would already have been reported a long time ago.
For now I think it might be nice to have a dialog box pop up announcing that Firefox is closing if it takes more than about 3-5 seconds. how about a counter in the dialog showing how many seconds it is taking to close?