This came up as a forums topic: http://forums.mozillazine.org/viewtopic.php?t=581611&sid=a41cc1a0f747334917655818cc85b686 For me, the printing, or the preparing to print doesn't take that much time, but I still get a very large spool file of appr. 240MB, while branch builds use appr. 40MB. Maybe, if you don't have that much memory in your computer, then it will a considerable longer time? This regressed between 2007-02-06 and 2007-02-07, I think a regression from bug 177805, somehow. I have only tested with philly.com, but according to the mozillazine topic there are more pages that suffer from this.

For comparison : HP Laserjet 4100 (PCL mode) 2.0.0.6 : 1.143KB latest trunk : 6.688KB HP Business Inkjet 2250 (PS mode) 2.0.0.6 : 2.534KB latest trunk : 7.443KB I have written PostScript before, I'll try to make sense of the difference. My first reaction (sorry I have to leave within a minute), is that the images are much larger. And that all text is send as images to the printer. Must be a side effect of the Cairo library (transparency, anti-aliasing, ...). There are no fonts included anymore.

Dupe of bug 362325 probably, although that one is only looking at performance (and originally filed for the Mac). But these big files might be one of the reasons that printing is slow.

Note that bug 362325 was filed before this regressed. I can clearly see this regressed between 2007-02-06 and 2007-02-07, where the latter dated build takes up >400MB as spool file (making my computer miserable). In current trunk builds, it's not as bad, just appr. 240MB, as I said.

how about bug 377336?

IMVHO it is actually bug 378307. Printing pages which doesn't make cairo to use image fallbacks is almost equally fast as with branch builds.

Ok, thanks, there are already quite a few bugs that handle the same issue (and have blocking1.9+). I'm duping it against bug 377336 for now.