bug:wasteful/slow method of processing code viewport

Showing 1-2 of 2 messages
bug:wasteful/slow method of processing code viewport Jim Michaels 1/29/13 8:48 PM
bug:wasteful/slow method of processing code viewport
new profile using firefox -p makes no difference.
fb 1.11.1
Pentium 4 HT 2.8GHz (1 core 2 thread)
firefox 18.0.1
XP Pro SP3 32-bit
2x2TB hard disks

disk free space (* next to drive is current/windows drive)
Tue 01/29/2013 19:23:50.81||E:\|>\u\df * -a
       C:---------------------------------******* 178.79GB/ 896.18GB( 19.95%)
       D:-----------------***********************   1.18TB/   2.00TB( 59.07%)
      *E:-----------------------------------***** 145.93GB/   1.00TB( 14.52%)
       F:-----------------------*****************   3.38GB/   7.75GB( 43.67%)
       G:------------------------****************   3.14GB/   7.51GB( 41.90%)
       H:------------------------****************   3.20GB/   7.56GB( 42.33%)
       I:-----------------------*****************   3.44GB/   7.80GB( 44.07%)
       J:-----------------------*****************   3.33GB/   7.69GB( 43.28%)
       K:-----------------------*****************   3.44GB/   7.81GB( 44.10%)
       L:-----------------------*****************   3.28GB/   7.66GB( 42.89%)
       M:--------------------------**************   2.91GB/   8.01GB( 36.41%)
     TOT:-------------------------***************   1.53TB/   3.96TB( 38.67%)
TOT USED:----------------************************   2.43TB/   3.96TB( 61.33%)
the small partitions are for swap/VM. 32-bit system has limit of 4GiB swap chunks, but you can have lots of them.

something critical to reproducing this problem MIGHT be:
- the slowness of my machine. Pentium 4 HT 2.8GHz (1 core 2 thread). it's equivalent to today's cheapest $300 laptop. really bad. I am used to it, but sometimes it seems really slow.
- my platform or OS. I have XP Pro SP3 32-bit

I would ask that you please try doing your testing on a really slow box like an old P4 xp machine. you might be able to repro the results. I have had similar troubles with fb before on this platform. this is all I have right now for a platform. I can also try on a windows 8 p4 box.

test case:

1.just updated (and fixed, to all appearances) my javascript code visit http://Jesusnjim.com/test/js-stringtodtime.html
2.turn on debugger
3.enable all panels
5.view the script time-funcs.js and hold down the scroll thumb (the little block) and scroll slowly. this code is about 2k lines of code. it should cause fb to sporadically make the code viewer go completely white. you can also try viewing jquery's source.
6.go to function NamedTimeStringToDTime and put breakpoints on every instance of GetIntegerKeywordPair() and StringDateTimeToDTime(). 174, 978, 1850,1851,1882,1883, 1937,1945,1949,1956,1958,1967,1975,1979,1989,1997, 2012,2020,2024,2043,
7.hit the refresh button on the browser. this will get the code going.
8.scroll the mouse wheel. the code view will clear the code view (go white), contract the font size and then go to a normal font size.  maybe one of these steps (contracting the font) is not necessary. maybe not even clearing/whiting the code view? I see this because I am on a slow box with big code.

efficient method of scrolling and viewing code in code viewport.

seems inefficient...
not sure clearing is necessary might be.
displaying the code twice doesn't look like it is necessary, so this looks like a time/cpu cycles waster (inefficient).

Re: bug:wasteful/slow method of processing code viewport Sebastian Zartner 2/5/13 2:50 PM
We're currently working on moving to a new Script panel display, with which slow scrolling should not be a problem anymore. We hope to have this ready for Firebug 1.12.