Re: [Flightgear-devel] Bones and IK
On Wednesday 16 June 2004 07:21, Erik Hofman wrote: > Lee Elliott wrote: > > Re the above, it would help if all the scenery could be made invisible > > because while rotating the view around the model to get the right angle > > to observe a particular animation I often find that the ground gets in > > the way and obscure the view. > > What really would help is a separate aircraft design program... > > Erik Ain't sayin' nuffin' ;) Actually, I love the modeller in the app I use - apparently every version of the s/w I use has been described as having a 'steep learning curve' but to be honest, I find it totally logical to the extent that I don't need to remember loads of stuff because it's just as quick to work it out each time I need to do it (not that I have on every model). Actually, it's easy enough too, to work out the exact anim axis by measuring them in whatever 3d app you're using - simple enough maths that even I can do it. Nope - the only two real problems are; the scenery getting in the way, and not being able to easily 'shift' the viewpoint i.e. to be able to get a view aligned with the central axis but off-set to the ailerons - by rotating you can only ever see them at an angle, other that that FG is ok, if with a bit of a heavy overhead:) LeeE ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] Bones and IK
On Thursday 17 June 2004 00:01, Lee Elliott wrote: > On Wednesday 16 June 2004 07:21, Erik Hofman wrote: > > Lee Elliott wrote: > > > Re the above, it would help if all the scenery could be made invisible > > > because while rotating the view around the model to get the right angle > > > to observe a particular animation I often find that the ground gets in > > > the way and obscure the view. > > > > What really would help is a separate aircraft design program... > > > > Erik > > Ain't sayin' nuffin' > > ;) > > Actually, I love the modeller in the app I use - apparently every version > of the s/w I use has been described as having a 'steep learning curve' but > to be honest, I find it totally logical to the extent that I don't need to > remember loads of stuff because it's just as quick to work it out each time > I need to do it (not that I have on every model). > > Actually, it's easy enough too, to work out the exact anim axis by > measuring them in whatever 3d app you're using - simple enough maths that > even I can do it. > > Nope - the only two real problems are; the scenery getting in the way, and > not being able to easily 'shift' the viewpoint i.e. to be able to get a > view aligned with the central axis but off-set to the ailerons - by > rotating you can only ever see them at an angle, other that that FG is ok, > if with a bit of a heavy overhead:) > > LeeE Oops! - that '(not that I have on every model)' should've been appended to the next paragraph - I blame the evils of drink. LeeE ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] Clickable 3d instruments
I have just tried it again. The path tag seem to expect a path to the XML file, so specifying a path to a model will result in an error and that stupid "glider". I have also assign the path to the XML file and just animate away. There is no error, but there is no movement in the model either. I then tried animating objects one at a time and all at once to find out if the two would make any difference. The result is still the same -- nothing. Regards, Ampere On June 15, 2004 11:01 pm, Ampere K. Hardraade wrote: > I have just tried it. It doesn't seem to work. > > Regards, > Ampere > > On June 15, 2004 08:02 pm, Jim Wilson wrote: > > Well actually... h... Right now we're referencing xml files in the > > "path" property of the model arrays ( tags). What happens when > > you add animation tags inside a tag and make the path reference > > the .ac file instead of another wrapper? I think that might work. > > > > Like this: > > > > > > /pathname/switch.ac > > > > animation definition for this particular switch... > > > > > > > > Repeat for each switch. > > > > Best, > > > > Jim > > > > > > ___ > > Flightgear-devel mailing list > > [EMAIL PROTECTED] > > http://mail.flightgear.org/mailman/listinfo/flightgear-devel > > ___ > Flightgear-devel mailing list > [EMAIL PROTECTED] > http://mail.flightgear.org/mailman/listinfo/flightgear-devel ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
[Flightgear-devel] perl script to make fgsd object libraries from a directory tree
#!/usr/bin/perl -w
# [EMAIL PROTECTED] 2004/6/15
# GNU GPL
my $path='/';
my $pwd='/';
my $base;
my $depth=0;
my $name;
my $nick;
my $tab;
my $find_args;
my $dir;
my @files;
# Initailise the leaf array. Here we say / currently contains 0 leaves.
my @leaf; $leaf[0]=0;
my @types = qw(ac 3ds ase dfx flt md2 obj vrml vrml1 atg);
$find_args=' -follow -type f -name "*.'.pop(@types).'"';
foreach $i (@types) {
$find_args.=" -or -name \"*.$i\"";
}
$base=$ENV{PWD} unless ($base=shift());
$base=~/^--?h/ and &Help();
chdir($base) or die "could not chdir to $base";
@files = `find * $find_args`;
@files = sort @files;
foreach $i (@files) {
$i=~s!^([^/].*)!/$1!;
}
print ''."\n";
print "\n";
print "\n";
print " Library\n";
foreach $path (@files) {
$path=~s/^\.//;
$name=$path;
$path=~s/(.*)\/.*/$1/;
chomp($path);
$name=~s/.*\/(.*)/$1/;
chomp($name);
$nick=$name;
$nick=~s/(.*)\..*/$1/;
($pwd, $depth, @leaf) = &GotoRightDir($path, $pwd, $depth, @leaf);
# Now assume $path eq $pwd
$tab = &Tab($depth);
print "$tab\n";
print "$tab ${nick}\n";
print "$tab ${base}${path}/${name}\n";
print "$tab\n";
$leaf[$depth]++;
}
# Clean up
&GotoRightDir($path, $pwd, $depth, @leaf);
print "\n";
exit;
sub GotoRightDir {
# Ascend or descecnd directories until $path eq $pwd
# Make appropriate STDOUT along the way
my ($path, $pwd, $depth, @leaf) = @_;
my $branch;
# Just in case
$path='/' unless $path;
while ("$path" ne "$pwd") {
if ($path=~/^$pwd/) {
# Descend to the next dir
# branch is the next directory in $path after $pwd
$branch=$path;
$branch=~s/$pwd//;
$branch=~s/^\/?([^\/]*).*/$1/;
# / is a special case
$pwd=~s!/$!!;
$pwd.="/$branch";
# $depth is still set to the parent's level
$tab = &Tab($depth);
print "${tab}\n";
print "$tab $branch\n";
# Keep track of parent's number of leaves
$leaf[$depth++]++;
# This is a new dir, and has no leaves yet
$leaf[$depth]=0;
} else {
# Ascend toward /
$pwd=~s![^/]*$!!;
$pwd=~s!^(/.*)/$!$1!;
# Set $depth to parent's level
$depth--;
$tab = &Tab($depth);
print "${tab}\n";
}
}
return $pwd, $depth, @leaf;
}
sub Tab {
# Generate an indentation based on the parent's level
my $depth = shift;
my $tab='';
for ($i=0; $i<=$depth; $i++) {
$tab.=' ';
}
return $tab;
}
sub Help {
print "usage: $0 [-h] [path]\n\n";
print "Builds an fgsd xml object library and outputs it to STDOUT,\n";
print "If a path is given the library will be built with that as the root,\n";
print "otherwise the pwd will be used as the root.\n";
print "[path] may contain symlinks to other directories.\n";
exit 0;
}
___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] Re: 64 bit compile speed
Erik Hofman wrote: > No, 64-bit is by default a bit slower. It's on very special > occasions that 64-bit outperforms 32-bit (due to data transfer > rate...) That's true in the general case, because the larger pointer size effectively reduces the size of the L1 cache. But for this particular architecture, there are 8 extra (i.e. twice as many) general purpose registers available in 64 bit mode. Most CPU benchmarks seem to show about a 10-20% speedup when run on an AMD64 in 64 bit mode. But compilation (especially FlightGear's final link, which is, heh, awfully thrashy) is very often I/O bound, especially on a laptop with a slow disk. I guess in this case any CPU speedup is offset by the extra I/O to write the larger binary and the difference ends up as noise. :) Andy ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
