Bugzilla

Bugzilla's favicon is 140KB, which is quite large relative to other favicons of popular sites.  Can we bring that size down by a factor of 10% or so?

For comparison, here's the favicon that gets used on my machine on various other popular sites:
- TaoBao' favicon is 20KB (dimensions: 64x64)
- Reddit's favicon is 4KB (dimensions: 64x64)
- Google's favicon is 8KB (dimensions: 32x32)
- Zoom's favicon is 8KB (dimensions: 32x32)
- Instagram's favicon is 4KB (dimensions: 32x32)

So: bugzilla's favicon is many times bigger (in terms of bytes) than all these sites' favicons (7x the TaoBao favicon size, 35x the Reddit/Instagram favicon size).


This causes problems because:
(1) In Firefox, favicons incur some amount of performance cost for each tab that's using them (for now at least)
(2) Firefox developers tend to have **many** copies of bugzilla open in various tabs (maybe on the order of hundreds), either as part of their daily work (as a sort of to-do list, or things to cycle back to,) or as part of temporary deep-dives.
(3) As a result of (1) and (2), certain operations in Firefox can take on the order of several seconds for heavy Bugzilla users, when they're expected to be ~instantaneous.

The per-tab costs (labeled (1) above) are worth investigating and addressing; but in the meantime we could also bring down the costs by an order of magnitude or so if we simply reduced the size of the favicon that we use for BMO.

Bugzilla's favicon is 140KB, which is quite large relative to other favicons of popular sites.  Can we bring that size down by a factor of 10 or so (to e.g. 14KB or therabouts)?

For comparison, here's the favicon that gets used on my machine on various other popular sites:
- TaoBao' favicon is 20KB (dimensions: 64x64)
- Reddit's favicon is 4KB (dimensions: 64x64)
- Google's favicon is 8KB (dimensions: 32x32)
- Zoom's favicon is 8KB (dimensions: 32x32)
- Instagram's favicon is 4KB (dimensions: 32x32)

So: bugzilla's favicon is many times bigger (in terms of bytes) than all these sites' favicons (7x the TaoBao favicon size, 35x the Reddit/Instagram favicon size).


This causes problems because:
(1) In Firefox, favicons incur some amount of performance cost for each tab that's using them (for now at least)
(2) Firefox developers tend to have **many** copies of bugzilla open in various tabs (maybe on the order of hundreds), either as part of their daily work (as a sort of to-do list, or things to cycle back to,) or as part of temporary deep-dives.
(3) As a result of (1) and (2), certain operations in Firefox can take on the order of several seconds for heavy Bugzilla users, when they're expected to be ~instantaneous.

The per-tab costs (labeled (1) above) are worth investigating and addressing; but in the meantime we could also bring down the costs by an order of magnitude or so if we simply reduced the size of the favicon that we use for BMO.