00:37.38 |
IriX64 |
4 hrs i should have a cassie 4.1.1
:) |
00:50.11 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/misc/win32-msvc7/mged/mged.vcproj: don't define DM_X,
should be only using DM_WGL now on Windows |
00:52.42 |
*** join/#brlcad ``Erik
(i=erik@c-69-250-155-85.hsd1.md.comcast.net) |
00:58.07 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/src/libdm/
(dm-generic.c dm_obj.c libdm.dsp query.c tcl.c): decouple DM_X from
the other display manager interfaces, clean up the preprocessor
logic so Windows does not need to declare it |
01:21.33 |
IriX64 |
how long do we wait before we can post after a
CIA-9 post? |
01:26.11 |
*** join/#brlcad PrezKennedy
(n=Apathy@c-69-250-236-100.hsd1.md.comcast.net) |
01:26.47 |
IriX64 |
reboot, back later. |
01:37.43 |
*** join/#brlcad IriX64
(n=IriX64@toronto-HSE-ppp4303834.sympatico.ca) |
01:39.54 |
IriX64 |
Did I get an answer? Time after CIA-9 posts
before mere mortals can post again? |
02:32.15 |
brlcad |
IriX64: don't understand the
question |
02:33.11 |
brlcad |
CIA merely posts a notification when a CVS
commit is performed to the source repository, so everyone knows
exactly who is doing what and when generally speaking |
02:35.40 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/configure.ac:
check for pipe() |
02:43.44 |
IriX64 |
just thought i was interefering with
development. |
02:44.14 |
IriX64 |
thought there was a protocol of some type
here. |
02:50.46 |
brlcad |
only when you go on your pasting
rampages |
02:52.20 |
brlcad |
if there's design or development discussions
going on, off-topic and just basic "chit-chat" chatter would be
frowned upon but not when we're like this |
02:52.40 |
IriX64 |
thanks for the explanation. |
02:52.42 |
brlcad |
relatively low energy in the channel, simple
to manage.. nobody has to wade through to get a word in |
02:53.07 |
IriX64 |
my diff is huge :) |
02:53.35 |
IriX64 |
aieeeee patch broke ;) |
02:55.09 |
IriX64 |
9 minuts? thought my system was faster than
that. |
02:55.21 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/src/mged/ (10
files): decouple DM_X from the other display manager interfaces,
clean up the preprocessor logic so Windows does not need to declare
DM_X but also to avoid using _WIN32 where possible as
well. |
02:55.50 |
brlcad |
you ever figure out why your wallclock counts
are so low? |
02:56.04 |
IriX64 |
didnt want to do the math. |
02:56.41 |
IriX64 |
lets see what bench has to say. |
02:59.05 |
brlcad |
your rtfm was fine |
02:59.16 |
brlcad |
but it is WAY off your wallclock count, which
isn't right |
02:59.57 |
IriX64 |
think i fixed it doing a raytrace right
now. |
03:00.08 |
brlcad |
implies something systemic is abnormal with
your system |
03:00.19 |
IriX64 |
hrmmph problem is bench is running
too. |
03:00.44 |
brlcad |
like it's abnormally busy doing way too many
other things while the benchmark is running, or you have several
benchmarks running simultaneously |
03:01.27 |
IriX64 |
when are you going to use your overlap tool to
fix the overlaps in havoc? |
03:02.58 |
IriX64 |
benching bldg391 whil i'm raytracing havoc,
get real. |
03:03.41 |
brlcad |
fixing havoc is left as an exercise for
someone in the community |
03:05.09 |
IriX64 |
496829 rays in 12.38 secs = 40147.80 rays/sec
(RTFM) and 496829 rays in 242.18 secs = 2051.49 rays/sec
(Wallclock) that better? |
03:05.33 |
IriX64 |
and bench just fixished. |
03:06.32 |
IriX64 |
2844 times faster than the 780. not bad i
guess. |
03:08.05 |
IriX64 |
now how the heck did translate break
(sigh) |
03:10.37 |
IriX64 |
what do you *mean its the shift key not the
ctrl key? (I'm in ramble on mode) |
03:11.16 |
IriX64 |
got them backwards thats all ill just change
the docs, cheap fix. :) |
03:11.17 |
brlcad |
no that's not better |
03:11.37 |
brlcad |
the rays/sec for wallclock shouldn't be more
than 10% different from the RTFM rays/sec |
03:11.54 |
IriX64 |
explain why. |
03:12.36 |
brlcad |
you're currently at about 2000%
different |
03:12.43 |
Twingy |
cause you aint running anything else |
03:12.50 |
brlcad |
i can't explain why without seeing what's
going on |
03:12.52 |
IriX64 |
but explain how its supposed to be
10% |
03:12.53 |
Twingy |
brlcad, news on parallels licenses? |
03:12.57 |
brlcad |
does it actually take 242 seconds to
raytrace? |
03:13.03 |
IriX64 |
yes |
03:13.17 |
brlcad |
Twingy: beats me, i haven't been in to the
office since well before siggy |
03:13.29 |
Twingy |
are you on vacation? |
03:13.47 |
brlcad |
not really |
03:13.50 |
brlcad |
just haven't been back |
03:13.56 |
brlcad |
it's only been a day |
03:14.16 |
Twingy |
but darlene has the order right? |
03:15.22 |
Twingy |
if not I'll have her order one for
me |
03:16.33 |
IriX64 |
perhaps itll help if we compare output
devices? |
03:17.08 |
brlcad |
i have no idea what the status is |
03:19.17 |
Twingy |
k, if Wendy comaplains I will get Drew to
order for me |
03:21.47 |
brlcad |
IriX64: 9 chances out of 10, it's not a
brl-cad configuration issue |
03:21.59 |
brlcad |
it's something going on with your
system |
03:22.10 |
brlcad |
run top or something to see what's eating up
cpu cycles |
03:25.15 |
IriX64 |
at the moment loadavg is 0% |
03:25.46 |
brlcad |
i find that hard to believe |
03:25.50 |
brlcad |
what does uptime say? |
03:25.57 |
IriX64 |
hey irc takes nothing. |
03:26.04 |
IriX64 |
nothing else going on. |
03:26.20 |
IriX64 |
let me check taskmanager. |
03:26.31 |
brlcad |
ooh, you're on windows? |
03:26.31 |
IriX64 |
both agree 0-1% load. |
03:26.42 |
IriX64 |
sorta |
03:26.56 |
brlcad |
softa? you either are or aren't :) |
03:27.07 |
IriX64 |
do a ver on me. |
03:27.27 |
brlcad |
that doesn't tell me anything about what
kernel/OS you're running |
03:27.27 |
IriX64 |
*nix high on windows. :) |
03:27.40 |
brlcad |
only you know that |
03:28.02 |
brlcad |
just by the fact that you have taskmanager
though pretty much says it |
03:28.07 |
IriX64 |
all right windows xp base running
cygwin_nt |
03:28.15 |
brlcad |
fair enough |
03:28.21 |
Twingy |
and your nickname is Irix64 *boggle* |
03:28.29 |
IriX64 |
you picked it. |
03:28.37 |
brlcad |
i suspect you do have hidden processes running
that are hiding themselves from the taskmanager |
03:28.47 |
brlcad |
rather cpu-intensive ones |
03:28.47 |
IriX64 |
no way |
03:28.56 |
IriX64 |
i would know |
03:28.58 |
brlcad |
heh |
03:29.00 |
IriX64 |
err |
03:29.06 |
brlcad |
and how praytell would you know? |
03:29.13 |
IriX64 |
yah groaner :) |
03:29.22 |
brlcad |
you have a massive indicator right there with
the performance number |
03:29.28 |
brlcad |
massive massive descrepancy |
03:29.44 |
IriX64 |
on my system *everything that runs is given a
name and entered into task manager |
03:29.45 |
brlcad |
tis the nature of much spyware |
03:30.03 |
IriX64 |
no empahtically no. |
03:30.04 |
brlcad |
nah, you can run apps that are hidden from
taskmanager |
03:30.10 |
brlcad |
used to do that back in the day |
03:30.12 |
IriX64 |
not *here. |
03:30.21 |
Twingy |
I think you should install parallels rather
than arguing with IriX64 |
03:30.30 |
brlcad |
you're on windows, it's possible |
03:30.50 |
IriX64 |
what are parallels? |
03:31.06 |
IriX64 |
ok brlcad its possible. |
03:31.08 |
brlcad |
heck, it could even be a virus infected dll or
other system-level injection |
03:31.21 |
brlcad |
that could go completely undetected |
03:31.29 |
IriX64 |
what do *you raytrace to? |
03:31.41 |
brlcad |
parallels is a virtual machine |
03:31.52 |
IriX64 |
vmware style? |
03:31.52 |
brlcad |
sort of like vmware |
03:31.52 |
brlcad |
but cheaper |
03:32.01 |
brlcad |
less developer-feature-filled |
03:32.11 |
IriX64 |
got os/2 warp 4 to run on vmware but what a
slug. |
03:32.12 |
brlcad |
but capable of running multiple simultaneous
OS images |
03:33.33 |
IriX64 |
i'm told brlcad can bench press three of
``Erik :) |
03:35.12 |
brlcad |
not from the looks of his belly recently :)
*ahem* |
03:44.49 |
IriX64 |
you've seen my belly? *shrug* I don't hide
it. |
03:45.00 |
IriX64 |
:) |
03:45.44 |
IriX64 |
you're wondering about me lets get peronal,
I'm Mario and i'm class of 54. |
03:46.07 |
IriX64 |
errrr personal too. :) |
03:46.57 |
brlcad |
actually i meant erik's but good to know
mario |
03:47.06 |
IriX64 |
brlcad i put you in your early 40's? |
03:47.32 |
brlcad |
nah, i'm a young whipper snapper |
03:47.36 |
brlcad |
at least that's what I tell myself |
03:47.47 |
IriX64 |
hah so do I.:) |
03:50.10 |
*** join/#brlcad DTRemenak
(n=Daniel_R@c-24-23-59-104.hsd1.ca.comcast.net) |
03:51.45 |
*** join/#brlcad IriX64
(n=mario_du@toronto-HSE-ppp4303834.sympatico.ca) |
03:52.08 |
IriX64_ |
my night at the opera, see if you can find
me. |
04:08.29 |
*** join/#brlcad IriX64
(n=IriX64@toronto-HSE-ppp4303834.sympatico.ca) |
04:08.47 |
IriX64 |
back to shift grips :) |
04:16.08 |
*** join/#brlcad IriX64
(n=Who@toronto-HSE-ppp4303834.sympatico.ca) |
04:18.14 |
IriX64 |
IriX64_: Doofus you said shift grips
:) |
04:25.38 |
IriX64 |
we were talking about rtfm to wallclock times,
whats the ratio? |
04:26.53 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/BUGS: debugbu 2
immediately reports a bu_vls_free() error.. apparently been a
problem since 4.5 days at least (bug found in
doc/html/manuals/mged/bugs) |
04:27.47 |
IriX64 |
bug i can bring up two instances of mged (or
is this by design?) |
04:28.18 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/doc/html/manuals/mged/ (Makefile.am bugs): get rid of
doc/html/manuals/mged/bugs .. the main bug in there was verified to
still be a problem and was added to the BUGS file. either way, we
don't need this old 'bugs' file any more. |
04:31.05 |
IriX64 |
may i send you my mged.exe? |
04:32.13 |
IriX64 |
you think only cvs people are capable of
contributing? |
04:34.13 |
brlcad |
the rtfm rays/sec and the wallclock rays/sec
should be within 10% of each otehr, so a ratio of .. 9:10 or
thereabouts? |
04:34.47 |
brlcad |
being able to run multiple mgeds is definitely
by design |
04:36.04 |
IriX64 |
9:10? lets define this, a ray is considered
finished when? |
04:37.18 |
brlcad |
there's documentation on the various values
that I'd rather not describe over irc .. maybe read the
benchmark.tr nroff document if you *really* want to know |
04:37.35 |
brlcad |
but suffice it to say that they should NOT be
that different rays/sec if everything is "normal" |
04:37.51 |
IriX64 |
*really* yes. |
04:39.49 |
brlcad |
rtfm takes into consideration several aspects
outside of your control like how long it takes a process to start
up only counting the CPU time that you are actually
allocated |
04:40.17 |
brlcad |
wallclock time counts the amount of actual
time elapsed as measured by a clock |
04:41.16 |
IriX64 |
how do you measure cpu time what function
please clock() or time()? |
04:41.20 |
brlcad |
those shouldn't only be within 10% .. they
should be nearly identical .. but some systems are better than
others and there's some minor variance based on how busy systems
are, how long the context switches are, etc |
04:41.53 |
IriX64 |
why dont i just look in the code :) |
04:43.47 |
brlcad |
with yours as big as it is, it's pointing a
big finger at there being something serious going on like something
hogging the cpu or a bug in the wallclock computation
code |
04:44.04 |
brlcad |
the latter isn't the case as you mentioned
that's about how long it takes |
04:44.32 |
brlcad |
which is incredible as that is very very slow
for the benchmark images |
04:44.43 |
brlcad |
they should just take a few seconds per
frame |
04:47.11 |
IriX64 |
err brlcad: thats the output of raytraceing
while actually rendering the image to the tube quite fast
considering what its doing. |
04:47.48 |
brlcad |
it should just take a few seconds to fully
generate the image |
04:48.03 |
IriX64 |
its a solid model brlcad think. |
04:49.11 |
IriX64 |
i should send you a screen shot of my screen
with one rendered, if i knew how i would. |
04:49.55 |
IriX64 |
be back in a moment my attention is required
outside. |
04:52.04 |
brlcad |
i never said it wasn't a solid model |
04:52.16 |
brlcad |
never said that I doubt it renders incorrectly
either |
04:52.31 |
brlcad |
s/incorrectly/correctly/ |
04:52.46 |
brlcad |
the only thing wrong is the performance.. that
is WAY too slow |
04:53.24 |
brlcad |
unless you're running XP on a 486 or something
slower |
04:59.27 |
*** join/#brlcad IriX64_
(n=IriX64@toronto-HSE-ppp4303658.sympatico.ca) |
05:02.04 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/doc/html/manuals/mged/ (mug mug_camo): get rid of the old
path |
05:03.23 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/doc/mged.tr: get
rid of the old path |
05:05.44 |
IriX64_ |
486??? |
05:05.55 |
IriX64_ |
they still make those? |
05:05.57 |
IriX64_ |
:) |
05:07.19 |
IriX64_ |
ever hear of the dos game doom? |
05:07.47 |
IriX64_ |
music and all, vmware suck my arse with your
face :) |
05:09.07 |
IriX64_ |
hairy virtualizing interrupt and dma and
sundry hardware the game expects. |
05:15.55 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/src/ (14 files in
7 dirs): get rid of the old 'cad' path
references/assumptions |
05:16.36 |
IriX64_ |
simple concept really, you provide a dos
environment and any dos program that expects a certain environment
usually probes irq's and dma's and such if you feel the probe map
it to whats really there if the program doesn't probe try to
provide "generic" hardware. |
05:30.08 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/libdm/dm-ogl_win.c: removed the replaced file.
dm-ogl_win.c is no more, replaced by the dm-wgl.c interface
file. |
05:42.11 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/libtclcad/libtclcad.dsp: take a manual stab at an
initial msvc build file for libtclcad (needed for the sketch
editor) |
05:49.10 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/libsysv/Makefile.am: make vers.c correctly rebuild if
a source file changes |
05:49.29 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/libtclcad/Makefile.am: add the new windows build file
to the dist |
05:51.42 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/misc/win32-msvc/brlcad.dsw: add libtclcad to the msvc
brlcad dll build workspace |
05:55.59 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/mged/attach.c: looks like bezier canvas support should
now be available for both windows studio build projects. libtclcad
should build sans tk sources now, try enabling it. |
05:57.18 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/mged/attach.c: remove apparently dead code,
'do_2nd_attach_prompt()' |
06:08.50 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/src/mged/ (dm-X.c
dm-ogl.c dm-wgl.c mged_dm.h): move the common_dm() decl over to
mged_dm.h |
06:16.16 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/src/mged/
(chgview.c cmd.c vrlink.c): enable the 'pov' command, getting rid
of the peculiar undocumented TCP_FILES define. move cmd_pov over to
chgview.c like the comment suggests. (note that the command guts
are over in librt) |
06:39.08 |
*** join/#brlcad DTRemenak
(n=DTRemena@c-24-23-59-104.hsd1.ca.comcast.net) |
06:46.47 |
*** join/#brlcad clock_
(i=clock@84-72-88-162.dclient.hispeed.ch) |
06:47.37 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/src/libfb/ (26
files): get rid of the libfb-specific _LOCAL_ macro, instead using
HIDDEN like everyone else (currently still provided by
machine.h) |
08:23.25 |
*** join/#brlcad clock_
(n=clock@zux221-122-143.adsl.green.ch) |
13:42.51 |
*** join/#brlcad digitalfredy
(n=digitalf@200.71.62.161) |
16:37.43 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/tclscripts/mged/callbacks.tcl: quell warnings about
mged_display not being set, make sure the variables even
exist |
16:44.11 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/mged/edsol.c: |
16:44.11 |
CIA-9 |
BRL-CAD: add a lot of checks for valid
pointers as part of isolating a 'P' binding crash |
16:44.11 |
CIA-9 |
BRL-CAD: bug (sf bug 1375751: P shuts down
mged). the crash was the result of a bus |
16:44.11 |
CIA-9 |
BRL-CAD: error inside of Tcl_AppendResult of
all places, where the second |
16:44.11 |
CIA-9 |
BRL-CAD: Tcl_AppendResult() after the one
inside not_state() would cause a crash. this |
16:44.14 |
CIA-9 |
BRL-CAD: may indicate some other interp
initialization problem, but the fix in place does |
16:44.16 |
CIA-9 |
BRL-CAD: seem to keep things going
nicely. |
16:56.18 |
``Erik |
hm |
16:58.11 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/mged/chgtree.c: not_state() now reports the desired
state too, so no need for the extra printing |
16:59.06 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/mged/buttons.c: print out the expected/desired state
so the user has a clue as to what to do to resolve the problem when
not in the right editing mode. |
17:00.16 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/src/mged/edsol.c:
not_state now reports the desired state, so no need to print a
message. be consistent with the printing and the command name
category (sedit). |
17:00.36 |
*** join/#brlcad DTRemenak|RDP
(n=DTRemena@c-24-23-59-104.hsd1.mn.comcast.net) |
17:02.03 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/NEWS: fixed mged
crash on P binding when not in edit mode. this fixes sf bug 1375751
(P shuts down mged) reported by bob2. |
18:02.52 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/ (include/light.h
src/liboptical/sh_light.c): incremental update this time.. change
MAX_LIGHT_SAMPLES to SOME_LIGHT_SAMPLES |
18:07.20 |
*** join/#brlcad IriX64
(n=IriX64@toronto-HSE-ppp4303658.sympatico.ca) |
18:08.15 |
IriX64 |
what was that about a pencil tapping your
teeth? |
18:08.42 |
IriX64 |
easy,easy,money and your chicks for free,
right fatso :) |
18:09.49 |
IriX64 |
the doc says it's good for you ;) |
18:10.59 |
IriX64 |
should look that formula up, make some easy
money, only how do i get Canada to tap their teeth with a lead
pencil? |
18:13.53 |
IriX64 |
Captains mess, anyone ;) |
18:14.45 |
IriX64 |
mater and pater are here, be back in a
bit. |
18:23.03 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/misc/Makefile.defs: be sure to delete static and other
'regular' or noinst libraries during a make noprod |
18:24.10 |
IriX64 |
visit successfull :) |
18:29.01 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/ (include/light.h
src/liboptical/sh_light.c): next step in the incremental update,
dynamically allocate the light sample points array instead of using
a fixed size. |
18:33.47 |
IriX64 |
brlcad: i still can't get this math right,
2.61 secs RTFM 56.97 secs wallclock. |
18:34.18 |
IriX64 |
havoc.g |
18:35.13 |
IriX64 |
but its consistent across runs. |
18:37.14 |
IriX64 |
lets try bldg391. |
18:38.39 |
IriX64 |
1.25 rtfm 11.91 wallclock |
18:39.37 |
IriX64 |
lets try the deuce and a 1/2 |
18:51.10 |
IriX64 |
5.56 rtfm 93.55 wallclock |
18:52.13 |
IriX64 |
whats your e-mail, ill send you a .jpg shot of
my screen. |
18:54.01 |
IriX64 |
pastebin doesn't allow jpgs. |
18:54.10 |
brlcad |
jpg of your screen isn't useful |
18:54.24 |
IriX64 |
what do you need to help out here? |
18:55.14 |
brlcad |
I'd start by installing some antivirus and
antispyware software and see what all comes up |
18:55.26 |
IriX64 |
all right. |
18:56.29 |
brlcad |
clamAV is free, not sure how good their
antispyware is though |
18:56.45 |
IriX64 |
truck looks good though, nice green field its
sitting on. |
18:57.43 |
IriX64 |
gold tipped exhaust pipe... nothings to good
for the military. :) |
18:57.45 |
brlcad |
this looks good: http://www.spywareterminator.com/ |
18:58.07 |
IriX64 |
i'll try mcaffee first. |
18:59.08 |
brlcad |
mcafee sells a separate anti-spyware product
if all you have is their antivirus |
18:59.28 |
brlcad |
virus will be good to check for .. but should
be sure to check for both |
18:59.33 |
IriX64 |
think i bought the full package, ill dig it
out and see. |
19:15.24 |
*** join/#brlcad IriX64
(n=IriX64@toronto-HSE-ppp4303658.sympatico.ca) |
19:19.08 |
IriX64 |
look brlcad: it takes time to draw an image on
the screen, theres no way your 10% figure can apply. |
19:22.58 |
IriX64 |
ill either dcc send you a screenshot with the
figures on it or ill e-mail it to you. |
19:23.09 |
IriX64 |
and the picture on it. |
19:24.23 |
IriX64 |
e-mail address? |
19:24.55 |
IriX64 |
everythings clearly visible. |
19:25.23 |
brlcad |
if it takes 20x the ray-trace time, it would
imply a serious bug |
19:25.40 |
brlcad |
and I've run ray-traces on windows and don't
see that sort of different |
19:26.17 |
IriX64 |
all right how long does *yours take to draw
the m35 on your screen, im talking about the solid model picture,
not the wire frame. |
19:26.46 |
IriX64 |
err pardon? windows? |
19:26.47 |
brlcad |
here's what moss looks like for me: 759710.95
rays/sec (RTFM) 748857.07 rays/sec (wallclock) |
19:27.11 |
IriX64 |
it renders that fast? |
19:27.29 |
IriX64 |
im not talking make benchmark here. |
19:27.47 |
IriX64 |
thats the shot summary im getting my numbers
from. |
19:28.11 |
brlcad |
this has nothing to do with the model or
benchmark |
19:28.25 |
brlcad |
it's just relative performance
numbers |
19:28.32 |
brlcad |
doesn't matter what model either |
19:28.56 |
IriX64 |
screw its accuarate i used a stop watch. off
by maybe a second or two but its accyrate. |
19:29.06 |
IriX64 |
accurate too. |
19:29.24 |
brlcad |
sure, it's accurate |
19:29.27 |
IriX64 |
the math, she's good :) |
19:29.39 |
brlcad |
it's not once been a question to ME whether
the values were being computed correctly |
19:29.45 |
brlcad |
this is pretty stable code |
19:30.09 |
IriX64 |
*shrug* my code becomes stable with age too
:) |
19:30.25 |
brlcad |
which is exactly why it points at something
wrong with your system, why must I keep repeating this??
:) |
19:30.38 |
brlcad |
i personally don't really care |
19:30.42 |
brlcad |
it's your system |
19:30.49 |
brlcad |
but it does indicate some serious
problem |
19:31.03 |
IriX64 |
all right send me the pix of moss
(mario.dulisse2@sympatico.ca) |
19:31.04 |
brlcad |
not a problem with brl-cad or how the
ray-trace runs |
19:31.14 |
brlcad |
or what the results look like |
19:31.19 |
brlcad |
that is ALL fine |
19:31.26 |
brlcad |
what is wrong is how slow/fast it's running
for you |
19:31.51 |
brlcad |
you're not going to SEE anything related to
this problem with a pix |
19:31.59 |
brlcad |
other than it rendering slow |
19:32.13 |
brlcad |
is this not clear?? |
19:32.18 |
IriX64 |
you expect the pix to just appear? |
19:32.36 |
IriX64 |
drawing an image on the tube can not be that
fast. |
19:33.10 |
brlcad |
some images are pretty much immediate, others
not more than a few seconds |
19:33.19 |
brlcad |
you're looking at several minutes on your
system |
19:34.17 |
brlcad |
you mentioned m35 rendering - how long does it
take m35 to raytrace into a 512x512 window? |
19:34.39 |
IriX64 |
yes anyway im happy thanks, ill get that
mcaffee dug out. lost everything in a crash 3 months ago and am
slow to getting around to restoring some things. |
19:35.18 |
brlcad |
*shrug* .. if it is systemic, it means
something else basically has control of your system :) |
19:36.22 |
IriX64 |
512x512 just a sec |
19:36.47 |
brlcad |
you mentioned cygwin, that could be causing a
big performance penalty of some sort -- haven't looked at the
numbers there in a while |
19:37.01 |
IriX64 |
65.17 wallclock and 3.84 rtfm. |
19:37.50 |
brlcad |
see -- that means it only spent 4 seconds on
the cpu .. but for "some" reason more than a minute elapsed from
the time the ray-trace began until it completed |
19:38.21 |
brlcad |
there wasn't more than 3.84 seconds worth of
work, but something slowed it down |
19:38.29 |
IriX64 |
of course because the pixels have to be lit or
turned off sheesh. |
19:38.39 |
brlcad |
that could be a virus, spyware, cygwin..
something else |
19:39.03 |
brlcad |
i'd be surprised if cygwin was _that_ bad at
I/O |
19:39.29 |
IriX64 |
you using a hardware x? |
19:40.49 |
IriX64 |
rember there two layers of graphics here
windows and xserver althought the screens are seamless theres still
graphic overhead to contend with. |
19:43.08 |
IriX64 |
need a smoke be back shortly. |
19:43.11 |
brlcad |
there's not any configuration setup that
should have a discrepancy as large as yours is :) |
19:46.23 |
``Erik |
*yawn* |
19:47.43 |
``Erik |
heh, sean, I did an update and now libfb
doesn't want to compile, it's seeing two do_event()s defined, and
HIDDEN is reducing to /***/ instead of static o.O what'd you do tot
he thing? |
19:48.26 |
``Erik |
<-- hacked his if_X.c if_wgl.c and if_glx.c
to change the func name to like do_X_event() or do_wgl_event
depending on file.. think that would be worth a commit? |
19:48.46 |
brlcad |
``Erik: yeah, I know .. i have a fix for that
here |
19:49.38 |
brlcad |
there's a few of them |
19:56.00 |
*** join/#brlcad dtidrow_work
(n=dtidrow@host169.objectsciences.com) |
20:02.35 |
IriX64 |
brlcad: you are a very patient man, thanks for
the input. |
20:14.55 |
*** join/#brlcad clock_
(i=clock@84-72-93-187.dclient.hispeed.ch) |
20:47.52 |
*** join/#brlcad mg`
(n=mathieu@ALille-151-1-60-91.w83-198.abo.wanadoo.fr) |
20:54.37 |
*** join/#brlcad _mathieu
(n=mathieu@ALille-151-1-40-42.w83-198.abo.wanadoo.fr) |
21:17.58 |
*** join/#brlcad mathieu_
(n=mathieu@ALille-151-1-2-169.w82-127.abo.wanadoo.fr) |
21:20.12 |
*** join/#brlcad mathieu__
(n=mathieu@ALille-151-1-35-154.w83-198.abo.wanadoo.fr) |
21:33.52 |
*** join/#brlcad _mathieu
(n=mathieu@ALille-151-1-70-209.w86-207.abo.wanadoo.fr) |
21:46.07 |
*** join/#brlcad mathieu_
(n=mathieu@ALille-151-1-23-29.w82-127.abo.wanadoo.fr) |
21:50.19 |
*** join/#brlcad digitalfredy
(n=digitalf@200.71.62.161) |
21:53.01 |
*** join/#brlcad mathieu__
(n=mathieu@ALille-151-1-43-79.w83-192.abo.wanadoo.fr) |
21:53.38 |
*** part/#brlcad digitalfredy
(n=digitalf@200.71.62.161) |
22:07.19 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/liboptical/sh_light.c: final mod to allow for
arbitrary counts of light point samples. the point sample array is
allocated in batches of SOME_LIGHT_SAMPLES. |
22:16.46 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/liboptical/sh_stack.c: avoid debug mode namespace
conflicts with libfb's stack interface (where HIDDEN becomes
/**/) |
22:16.56 |
CIA-9 |
BRL-CAD: 03brlcad *
10brlcad/src/liboptical/sh_null.c: avoid debug mode namespace
conflicts with libfb's null interface (where HIDDEN becomes
/**/) |
22:17.48 |
``Erik |
sean, you doing the 'week off after siggraph'
thing? |
22:18.44 |
dtidrow_work |
c'mon, SIGGRAPH _is_ a week off ;-) |
22:18.59 |
dtidrow_work |
'cept for the parties... |
22:19.03 |
``Erik |
heh |
22:19.09 |
``Erik |
yeah, last year, I hurt my ankle pretty
bad |
22:19.22 |
``Erik |
falling down, uh, stairs, at a nightclub, for
an afterparty |
22:19.22 |
``Erik |
heh |
22:19.28 |
``Erik |
<-- retarded |
22:19.30 |
dtidrow_work |
which one? |
22:19.51 |
``Erik |
hrm? which club? or which party? |
22:19.58 |
``Erik |
it was at 'the mayan' in downtown la |
22:19.59 |
dtidrow_work |
which party |
22:20.08 |
``Erik |
I don't remember who was throwing it |
22:20.19 |
``Erik |
some flyers were handed out |
22:20.22 |
``Erik |
like the acm one or something |
22:20.25 |
dtidrow_work |
that's where the chapters party was,
right? |
22:20.32 |
``Erik |
yeah, I think that was it |
22:20.35 |
``Erik |
that sounds familiar |
22:20.37 |
dtidrow_work |
k |
22:20.43 |
dtidrow_work |
was there too :-) |
22:20.51 |
``Erik |
with the stirppers dancing on the
boxes? |
22:20.54 |
``Erik |
'cept they didn't strip |
22:20.55 |
``Erik |
heh |
22:21.02 |
dtidrow_work |
yeah |
22:21.07 |
``Erik |
and the big circular thing they were playing
crap on |
22:21.15 |
``Erik |
<-- stayed there pretty late |
22:21.17 |
dtidrow_work |
we nicknamed them the 'anime dancers'
:-) |
22:21.27 |
``Erik |
y'know the ampitheater seating on the upper
level? |
22:21.31 |
dtidrow_work |
yep |
22:21.41 |
``Erik |
I was walking down those, hit a corner of a
stemp with my heal trying to get into the seating |
22:21.47 |
dtidrow_work |
ouch |
22:21.54 |
``Erik |
black on black, smoke, dark, vodka... bad
mojo |
22:22.07 |
dtidrow_work |
yep |
22:22.17 |
``Erik |
came down on my ankle and butt, was hobbling
out, coulnd't move it the next morning |
22:22.35 |
``Erik |
when I got a 5am call to fix a diskless
cluster |
22:22.37 |
dtidrow_work |
was awfully expensive this year - $7 for a Sam
Adams |
22:22.46 |
``Erik |
ew |
22:22.47 |
dtidrow_work |
that's one bottle |
22:22.59 |
``Erik |
hah, go down the street and buy a 6 for less,
hah |
22:23.13 |
dtidrow_work |
I can get a six-pack of them for
$7.50 |
22:23.41 |
``Erik |
huh, I've seen sam cheaper |
22:23.43 |
dtidrow_work |
was sorely tempted to go on the brewery tour,
as the brewery was only about 5-6 miles away from my
hotel |
22:24.02 |
``Erik |
heh |
22:24.08 |
``Erik |
this year? |
22:24.12 |
dtidrow_work |
yep |
22:24.18 |
dtidrow_work |
Boston |
22:24.30 |
``Erik |
I wanted to go, but there was a clamp put down
on all travel that I couldn't squirm out from under |
22:24.36 |
dtidrow_work |
ick |
22:24.36 |
``Erik |
so I had to call in, cancel my hotel
reservation, etc |
22:24.42 |
``Erik |
sucked |
22:24.50 |
dtidrow_work |
indeed |
22:25.06 |
``Erik |
now they're trying to send me to utah for some
realtime raytracing convention |
22:25.44 |
dtidrow_work |
there were several papers about that this
year, or maybe courses |
22:26.17 |
``Erik |
there have been for a while |
22:26.20 |
``Erik |
and all the same players, heh |
22:26.27 |
``Erik |
slusallek, etc |
22:29.20 |
dtidrow_work |
lol |
22:35.18 |
dtidrow_work |
guess he's totally distracted |
22:36.24 |
``Erik |
he must have a sexy piece of code keeping him
distracted |
22:37.05 |
dtidrow_work |
heh |
22:47.52 |
CIA-9 |
BRL-CAD: 03brlcad * 10brlcad/src/libfb/
(if_X.c if_ogl.c if_wgl.c): avoid debug mode namespace conflicts
when multiple interfaces are compiled and HIDDEN becomes
/**/) |
22:49.09 |
brlcad |
mm.. sexy code indeed |
22:49.52 |
brlcad |
``Erik: not the entire week |
22:50.10 |
brlcad |
i'm on a coding roll on many fronts |
22:50.22 |
``Erik |
ah, I was hoping in your aid in something... I
can wait if you'd rather not be arsed with my issues :) |
22:50.30 |
brlcad |
it'll be all over once I go in, so I'm going
with the flow |
22:51.37 |
brlcad |
feel free to ask, if it's too complicated I'll
defer, but not likely |
22:51.37 |
``Erik |
(mem leak in my shot routine, I not grokin'
something) |
22:51.59 |
brlcad |
set bu_debug to 2 |
22:52.39 |
brlcad |
basic bounds checks, won't catch the leak of
course, but it's a start |
22:54.28 |
``Erik |
that's a lot of info |
22:55.48 |
``Erik |
nothing about allocations and deallocations,
though |
23:11.28 |
brlcad |
hmm.. shouldn't be "a lot" .. note that one
flag is for rt_debug, another for bn, and another for bu |
23:12.20 |
brlcad |
i mean, it'll be a lot of messages .. the ones
that are provided during bu_malloc/bu_free/etc |
23:32.35 |
IriX64 |
try my_free(char *ptr)
{free(ptr);ptr=NULL;} |
23:34.40 |
IriX64 |
or my_free(char
*ptr){if(ptr){free(ptr);ptr=NULL}} |
23:38.11 |
IriX64 |
rtshot or show shot ``Erik? |
23:41.56 |
IriX64 |
in rtshot you have two allocs but no free's.
kosher? |
23:47.36 |
IriX64 |
sigh my memory leaks constantly, silicon might
help ;) |
23:49.28 |
IriX64 |
err my_free(char*
ptr){if(ptr){free(ptr);ptr=NULL;return(ptr)}}return; |
23:49.44 |
IriX64 |
been awhile :) |
23:50.13 |
IriX64 |
err my_free(char*
ptr){if(ptr){free(ptr);ptr=NULL;return(ptr)}}return(ptr); |
23:51.44 |
IriX64 |
make distclean |
23:52.03 |
IriX64 |
should have auto switching windows...
sorry. |