O(n) vs O(bigger than I wanted) RT @bz_moz: Memo for bug filers: "memory leak" and "memory usage" are not the same thing.
1
@littlecalculist @bz_moz Given how hard it is for _us_ to get enough insight to differentiate such cases, maybe forgive the bug filers :)
2
@graydon_moz @bz_moz True... in fact I honestly have no idea how to diagnose memory consumption problems.
1
@littlecalculist @graydon_moz It starts depending on application. But as a first cut, stable memory usage == "not leak". With caveats. ;)
1
Replying to @really_bz
@bz_moz @littlecalculist @graydon_moz In Fx, I'd say caching/fragmentation is ok; doing X a second time should peak no higher than first.

Sep 4, 2012 · 12:58 AM UTC

4
Replying to @davidbaron
@davidbaron @bz_moz @littlecalculist There's fast growth and slow growth. It's a complex program. Still 2x resident after 24h, for me.
2
@graydon_moz @bz_moz @littlecalculist Does it keep increasing? I had Firefox running for 48 days (ending last week), quite tolerably.
3
more replies
Replying to @davidbaron
@davidbaron @bz_moz @littlecalculist Growth in heap-unclassified %age over long run. Despite formidable memshrink heroics.
1
Replying to @davidbaron
@davidbaron @bz_moz @littlecalculist Don't get me wrong, it's a hard-hard problem, and we're on a solid positive trend.
Replying to @davidbaron
@davidbaron @bz_moz @littlecalculist Unlike when I worked on it, and kept looking for cheap/quick fixes. Seems it's a million papercuts.