 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hi,
Think this is a bug, has happened three times (across reboots of machine
etc etc). Causes rendering to abort with a bogus (and insulting) error.
POV says
this:
Parsing.....................................................................
...
[snip].................................................
.................Could not find file 'scene_shell.pov'
#declare yy = yy + incrementy;
#local indexy = indexy + 1;
#end (yy <= maxy)
#local zz = domath(xx,yy) <----ERROR
Tesseract:Apps:Multimedia:3D Apps:renderers:POV-Ray
3:Include:math2.inc:144: error: Cannot open macro file scene_shell.pov.
Check that the file is in a directory specifed with a +L switch
or 'Library_Path=' .INI item. Standard include files are in the
include directory or folder. Please read your documentation carefully.
'scene_shell.pov' is the scene file (funnily enough), domath is defined in
scene_shell and then again (if it isn't defined) in math2.inc.
[Yes, I've read the manual, and the 3.0, and the 2.2 and the 1.0 manuals
:-) I'm just a beginner, but I know where my towel is. math2.inc *is*, as
shown in the path above, in the standard include directory]
What caused the above (18 minutes into the parse - a big (to me at least)
smooth triangle mesh) was the insertion of an audio CD. Tried it again
(after rebooting, just in case an OS gremlin stabbed POV in the back) -
same result, and again (data CD this time) same result. Interestingly a
floppy doesn't cause the problem (locked or unlocked). Haven't tried other
storage media...
It *always* dies on the same line, hence I think it has something to do
with having the same macro declared in both files. The only other includes
are 'standard' includes.
[I only read C at a snail's pace and haven't even glanced at the source]
Now there's a small chance an extension to my OS is causing the problem,
but I don't have anything running that does anything on insertion of a disk
other than the bult-in audio CD player... which shouldn't affect data CDs.
Machine is a PPC7500 upgraded to a G3@312MHz (512k of backside cache
@124MHz). POV has 27MB of memory allocated and uses about 2/3 of that when
rendering the scene (yes, the scene parses if I don't insert a CD).
So, can anyone else confirm?
What about on 68k machines?
Have Fun
Martin
--
Owner/Operator - Tesseract Computing
<hypercube 'at' tesseract.com.au> or just reply.
Computer Systems Officer - Tourism Tasmania
<Martin.Crisp 'at' tourism.tas.gov.au>
I speak for me.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <B3B### [at] 147 10983.2> , "Martin Crisp"
<val### [at] tesseractcomau> wrote:
> Hi,
>
> Think this is a bug, has happened three times (across reboots of machine
> etc etc). Causes rendering to abort with a bogus (and insulting) error.
> POV says
> this:
>
> Parsing.....................................................................
> ...
> [snip].................................................
> .................Could not find file 'scene_shell.pov'
>
> #declare yy = yy + incrementy;
> #local indexy = indexy + 1;
> #end (yy <= maxy)
> #local zz = domath(xx,yy) <----ERROR
>
> Tesseract:Apps:Multimedia:3D Apps:renderers:POV-Ray
> 3:Include:math2.inc:144: error: Cannot open macro file scene_shell.pov.
> Check that the file is in a directory specifed with a +L switch
> or 'Library_Path=' .INI item. Standard include files are in the
> include directory or folder. Please read your documentation carefully.
>
> 'scene_shell.pov' is the scene file (funnily enough), domath is defined in
> scene_shell and then again (if it isn't defined) in math2.inc.
> [Yes, I've read the manual, and the 3.0, and the 2.2 and the 1.0 manuals
> :-) I'm just a beginner, but I know where my towel is. math2.inc *is*, as
> shown in the path above, in the standard include directory]
>
> What caused the above (18 minutes into the parse - a big (to me at least)
> smooth triangle mesh) was the insertion of an audio CD. Tried it again
> (after rebooting, just in case an OS gremlin stabbed POV in the back) -
> same result, and again (data CD this time) same result. Interestingly a
> floppy doesn't cause the problem (locked or unlocked). Haven't tried other
> storage media...
>
> It *always* dies on the same line, hence I think it has something to do
> with having the same macro declared in both files. The only other includes
> are 'standard' includes.
>
> [I only read C at a snail's pace and haven't even glanced at the source]
>
> Now there's a small chance an extension to my OS is causing the problem,
> but I don't have anything running that does anything on insertion of a disk
> other than the bult-in audio CD player... which shouldn't affect data CDs.
>
> Machine is a PPC7500 upgraded to a G3@312MHz (512k of backside cache
> @124MHz). POV has 27MB of memory allocated and uses about 2/3 of that when
> rendering the scene (yes, the scene parses if I don't insert a CD).
This sounds very much like a low memory problem. How much memory is left for
POV-Ray? If I understand you right this bug does no appear when no CD is
inserted. Try increasing the POV-Ray memory size so at least 10% more than
needed. You may also want to try to start with minimal extension set and
most up-to-date CD drivers, and btw which OS version are you using?
Thorsten
____________________________________________________
Thorsten Froehlich
e-mail: mac### [at] povrayorg
I am a member of the POV-Ray Team.
Visit POV-Ray on the web: http://mac.povray.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thorsten Froehlich wrote in message <37932952@news.povray.org>...
>In article <B3B### [at] 14710983.2> , "Martin Crisp"
><val### [at] tesseractcomau> wrote:
>
>> Hi,
>>
>> Think this is a bug, has happened three times (across reboots of
machine
>> etc etc). Causes rendering to abort with a bogus (and insulting)
error.
>> POV says
>> this:
>>
>>
Parsing.................................................................
....
>> ...
>> [snip].................................................
>> .................Could not find file 'scene_shell.pov'
>>
>> #declare yy = yy + incrementy;
>> #local indexy = indexy + 1;
>> #end (yy <= maxy)
>> #local zz = domath(xx,yy) <----ERROR
>>
>> Tesseract:Apps:Multimedia:3D Apps:renderers:POV-Ray
>> 3:Include:math2.inc:144: error: Cannot open macro file
scene_shell.pov.
>> Check that the file is in a directory specifed with a +L switch
>> or 'Library_Path=' .INI item. Standard include files are in the
>> include directory or folder. Please read your documentation
carefully.
[snip]
>> Machine is a PPC7500 upgraded to a G3@312MHz (512k of backside cache
>> @124MHz). POV has 27MB of memory allocated and uses about 2/3 of that
when
>> rendering the scene (yes, the scene parses if I don't insert a CD).
>
>This sounds very much like a low memory problem. How much memory is
left for
>POV-Ray? If I understand you right this bug does no appear when no CD
is
>inserted. Try increasing the POV-Ray memory size so at least 10% more
than
>needed. You may also want to try to start with minimal extension set
and
>most up-to-date CD drivers, and btw which OS version are you using?
Machine is running 8.6. Latest drivers (though admittedly the CD drive
is a CDR...). The Mac has 64MB of physical memory with Virtual set to
80MB. The OS is using about 16MB, POV has 27MB (so 43 in use total, the
VM should not be an issue) allocated and the death of parsing occurs
whether the POV is currently using 3MB of that or 13MB of that. I've
been keeping an eye on memory since I haven't worked with meshes much
and was concerned 27MB may not be enough... this is happening after a
recent launch of POV i.e. launch, enter key to open the scene file I was
last using (navigation services support is *nice* guys!), command-R,
wait five seconds insert CD (POV using only 1/10 of allocated memory at
this point) after a second or 2 parsing dies as above.
I'll double check that I don't have both Apple's CD driver and the Toast
CD reader running (now at work...).
I've got ZoneRanger wandering around my HDs somewhere if you want more
detailed info on memory usage....
Basically I'm batting 1000 with the bug (whether it be POV or some other
gremlin), the only reason for using a large file (actually the file
isn't large, nor really complex, but math.inc takes a while to generate
a few hundred thousand smooth triangles) is so POV is still parsing when
I insert the CD (though I suppose a trivial but long #while loop would
discount memory being the problem). Again I still think it's related to
the doubly-declared macro given the line that always causes the problem.
I'll write a couple of trivial macros and test it that way...
Have Fun
Martin
--
<Spa### [at] tesseractcomau> is a valid address
If you prefer mnemonic addresses you can use:
<Martin 'at' tesseract.com.au> or <Martin.Crisp 'at' Tourism.tas.gov.au>
My opinions should not be taken as indicative of Tourism Tasmania's
opinions
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3793b8ba@news.povray.org> , "Martin Crisp"
<Spa### [at] tesseractcomau> wrote:
> Machine is running 8.6. Latest drivers (though admittedly the CD drive
> is a CDR...). The Mac has 64MB of physical memory with Virtual set to
> 80MB. The OS is using about 16MB, POV has 27MB (so 43 in use total, the
> VM should not be an issue) allocated and the death of parsing occurs
> whether the POV is currently using 3MB of that or 13MB of that. I've
> been keeping an eye on memory since I haven't worked with meshes much
> and was concerned 27MB may not be enough... this is happening after a
> recent launch of POV i.e. launch, enter key to open the scene file I was
> last using (navigation services support is *nice* guys!), command-R,
> wait five seconds insert CD (POV using only 1/10 of allocated memory at
> this point) after a second or 2 parsing dies as above.
>
> I'll double check that I don't have both Apple's CD driver and the Toast
> CD reader running (now at work...).
>
> I've got ZoneRanger wandering around my HDs somewhere if you want more
> detailed info on memory usage....
No, if it happens that early during parsing there shouldn't be a memory
problem.
> Basically I'm batting 1000 with the bug (whether it be POV or some other
> gremlin), the only reason for using a large file (actually the file
> isn't large, nor really complex, but math.inc takes a while to generate
> a few hundred thousand smooth triangles) is so POV is still parsing when
> I insert the CD (though I suppose a trivial but long #while loop would
> discount memory being the problem). Again I still think it's related to
> the doubly-declared macro given the line that always causes the problem.
> I'll write a couple of trivial macros and test it that way...
Hmm, this hole issue is very strange. Parsing has nothing to do with any
Macintosh specific code and there is no code that specifically interacts
with CDs or any other media. It might be that you have an invalid path set
to search for include files. The Mac OS sometimes has strange ideas on where
to search for files. Does the bug appear if the CD is inserted before you
start rendering?
It would be great if you could create a simple source file to reproduce the
problem. Please send the file along with a bug-report form (in the
documentation folder) to mac### [at] povrayorg.
Thorsten
____________________________________________________
Thorsten Froehlich
e-mail: mac### [at] povrayorg
I am a member of the POV-Ray Team.
Visit POV-Ray on the web: http://mac.povray.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Tue, 20 Jul 1999 5:34 PM, Thorsten Froehlich <mailto:tho### [at] trfde>
wrote:
>In article <3793b8ba@news.povray.org> , "Martin Crisp"
><Spa### [at] tesseractcomau> wrote:
<snip>
>> I insert the CD (though I suppose a trivial but long #while loop would
>> discount memory being the problem). Again I still think it's related to
>> the doubly-declared macro given the line that always causes the problem.
>> I'll write a couple of trivial macros and test it that way...
>
>Hmm, this hole issue is very strange. Parsing has nothing to do with any
>Macintosh specific code and there is no code that specifically interacts
>with CDs or any other media. It might be that you have an invalid path
set
Gotta be something happening when a disk insert event happens. Weird that
it seems to be CD specific though. Haven't had a chance to check the
extension set yet though... (could still be this end - easily even).
>to search for include files. The Mac OS sometimes has strange ideas on
where
OK, I'll move my prefs sideways (only one path is showing in the prefs)...
no, still dies (specified a second include path too).
However, chose a different execution path in the include this time
(unsmoothed triangle routine also calls the doubly-declared macro), again
it died on the line asking for that macro - this is beyond coincidence -
and it's the original scene file it can't find. Somewhere a vRefNum is
being lost (guessing - but it sure looks like it - he says blithely without
sighting the code :-) - hey it could even be apple's code that's got the
hole [blasphemy])
>to search for files. Does the bug appear if the CD is inserted before you
>start rendering?
No. Nor does it matter if I (for example) start playing an audio CD after
parsing has started (didn't expect it would since data CDs cause the same
behaviour).
>It would be great if you could create a simple source file to reproduce
the
>problem. Please send the file along with a bug-report form (in the
>documentation folder) to mac### [at] povrayorg.
Will play...
Have Fun
Martin
--
Owner/Operator - Tesseract Computing
<hypercube 'at' tesseract.com.au> or just reply.
Computer Systems Officer - Tourism Tasmania
<Martin.Crisp 'at' tourism.tas.gov.au>
I speak for me.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Martin Crisp wrote in message ...
>On Tue, 20 Jul 1999 5:34 PM, Thorsten Froehlich
<mailto:tho### [at] trfde>
>wrote:
[snip]
>>It would be great if you could create a simple source file to
reproduce
>the
>>problem. Please send the file along with a bug-report form (in the
>>documentation folder) to mac### [at] povrayorg.
>
>Will play...
OK, got a scene/include pair happening at about 1:00am last night (the
short-blind-blonde-pomme insisted it was bedtime at that point)
(has a few options for where the macros get called from (original defn
in scene, original defn in inc, or swapping between redeclarations in
each...))
Will forward it on tonight when I get home...
The 'bug' is still present with the example scene/include.
Reducing my extensions set to (8.6base - CD/DVD driver + Toast CD
Reader) made no difference (I could reduce further, but sorry if it
clashes with these then there's a problem somewhere...)
BTW: Yes I agree that this is along the lines of:
Patient: Doctor it hurts when I do this <demonstrates> Ow!
Doctor: Don't do that then.
But it'd be nice to have some soothing music (Deep Purple: Nobody's
Perfect) playing as POV reveals the mistakes in my latest attempt at a
scene, and be able to change CDs while parsing :-)
Have Fun
Martin
--
<Spa### [at] tesseractcomau> is a valid address
If you prefer mnemonic addresses you can use:
<Martin 'at' tesseract.com.au> or <Martin.Crisp 'at' Tourism.tas.gov.au>
My opinions should not be taken as indicative of Tourism Tasmania's
opinions
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Wed, 21 Jul 1999 11:57 AM, Martin Crisp
<mailto:Spa### [at] tesseractcomau> wrote:
>Martin Crisp wrote in message ...
>>On Tue, 20 Jul 1999 5:34 PM, Thorsten Froehlich
><mailto:tho### [at] trfde>
>>wrote:
>
>[snip]
>
>>>It would be great if you could create a simple source file to
>reproduce
>>the
>>>problem. Please send the file along with a bug-report form (in the
>>>documentation folder) to mac### [at] povrayorg.
>>
>>Will play...
>
>OK, got a scene/include pair happening at about 1:00am last night (the
>short-blind-blonde-pomme insisted it was bedtime at that point)
>(has a few options for where the macros get called from (original defn
>in scene, original defn in inc, or swapping between redeclarations in
>each...))
>
>Will forward it on tonight when I get home...
>
>The 'bug' is still present with the example scene/include.
However, while writing out the bug report form (I hope you guys like
verbose mode) I discovered another factor. The macro called from the
include does have to live in the scene file, but the bug appears to be
related to changing windows in some way.
>Reducing my extensions set to (8.6base - CD/DVD driver + Toast CD
>Reader) made no difference (I could reduce further, but sorry if it
>clashes with these then there's a problem somewhere...)
The odds on it being POV (the core code) appear to have lengthened
considerably, the odds on it being a gremlin in the interface code in the
MacOS version appear to have moved to odds-on marginally worse than even,
with Toast CD Reader extension fighting against Quicktime (4.0.1?)
Extensions for a very short odds second favourite position.
>But it'd be nice to have some soothing music (Deep Purple: Nobody's
>Perfect) playing as POV reveals the mistakes in my latest attempt at a
>scene, and be able to change CDs while parsing :-)
Now Playing: Yello - Stella (Oh Yeah! is on at the moment)
Have Fun
Martin
--
Owner/Operator - Tesseract Computing
<hypercube 'at' tesseract.com.au> or just reply.
Computer Systems Officer - Tourism Tasmania
<Martin.Crisp 'at' tourism.tas.gov.au>
I speak for me.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Watch out for the Toast extension.
Why?
Toast is notorious ( I use it ) for being incompatible. I thought everybody
knew that. Try FWB CD Toolkit yet?
How?
When Toast is activated, it "patches a trap" (probably several). That is, it
takes a call to a standard function in the Mac OS and patches some
customized code over it. When Toast is finished, it removes the patch.
POV-Ray is probably calling the same patched trap ( I can't imagine what it
is ) while the CD is mounted and the patch is still in place. Since the CD
runs full time the patch is never released until the CD is ejected.
Apple has strongly warned against "patching traps" for years but everyone
does it - it's really very common. It usually works just fine if the release
code is well written. In the case of the CD, there's no way to release the
patch without halting the CD software.
If you want to see what code is executing, use MacsBug. It will break into
console mode and show you what's executing at the time of the error. Then
you can look it up in Apple's Tech database. From there you will know which
Toolbox function is being patched. MacsBug and related instructions are
bundled with CodeWarrior Lite, available for free from the MetroWerks web
site.
Of course, knowing this won't fix it.
Maybe you could just use a Walkman?
steve
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Steve wrote in message <37963C7F.79A6E38B@puzzlecraft.com>...
>Watch out for the Toast extension.
>
>Why?
>
>Toast is notorious ( I use it ) for being incompatible. I thought
everybody
>knew that. Try FWB CD Toolkit yet?
Funny, I've set up a multi-user, networked, automated scanning & CD
burning system based around Toast, FaceSpan (an AppleScript-based
interface builder), PlayerPro and 4th Dimension... last time it crashed
was when a hard disk died...
FWB:
No, and no intention of doing so.
>How?
[snip, yeah , yeah ,yeah :-) Word5 & cc:Mail were good at patching the
same traps...]
>Maybe you could just use a Walkman?
Hmm, my Sony DiscMan has been dead for about 5 years
Think I'll just hack the CD/DVD driver so it recognises non-apple drives
instead :-)
It may well be the Toast Extension, but it still looks as though
something funky is happening with POV regarding window
activation/deactivation.
Have Fun
Martin
--
<Spa### [at] tesseractcomau> is a valid address
If you prefer mnemonic addresses you can use:
<Martin 'at' tesseract.com.au> or <Martin.Crisp 'at' Tourism.tas.gov.au>
My opinions should not be taken as indicative of Tourism Tasmania's
opinions
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
>
>
> Funny, I've set up a multi-user, networked, automated scanning & CD
> burning system based around Toast, FaceSpan (an AppleScript-based
> interface builder), PlayerPro and 4th Dimension... last time it crashed
> was when a hard disk died...
Our CD burning stations are dedicated. Same for the scanning department.
Last time I launched Quark with Toast Pro active it destroyed the SCSI
directory on the entire drive. Now I master with Toast and a minimal
extension set.
Another reason for this is the fact that CPU interrupts while mastering can
ruin the mastering session. I never do anything else while mastering a CD.
I can't imagine how the CD would ever come out ok if you burn it while
raytracing.
The Apple CD driver and FWB are pretty benign, unlike Toast. They are both
just for reading, not mastering. So when I want a little music, I just use
one of them.
Toast may not be the culprit but with it's long, well earned history of
conflicts I'd just use something that works rather than worrying about it.
With my previous horrible experiences with Toast I'm not even gonna try
duplicating your problem.
If you must, use a debugger to find out for sure what's going on. There
might be something goofy about the compile not necessarily related to
either the source code or the Mac Toolbox. Check the MetroWerks web site
for documented bugs in their compilers. I don't use MPW so I don't know how
well their bugs are documented. If it was compiled on a PC with a freeware
compiler, then it's a miracle POV-Ray will even run. You might even
discover a DVD related bug in the POV source code.
Also, System 8.6 has numerous bugs coming to light. Apparantly there were
big-time changes in the release and quite a bit of stable software is now
buggy under the new OS (Quark 3.32 for example). I can give you a web page
dedicated to Mac OS conflicts and bugs - you might not have an isolated
problem. Won't be the first time Apple has released a diseased OS on us (
remember System 7?).
So, the quick-fix is just use the Apple CD driver for music and reserve
Toast for mastering.
steve
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
