Discussion:
[CsMain] unknown escape sequence?
(too old to reply)
Paul Garceau
2000-03-15 07:46:44 UTC
Permalink
Hi folks,

Not sure what is causing this. Looks like something having to
do with <eol> stuff. Any ideas?

c++ -c -o out/NT4/INTEL/optimize/sector.obj
libs/csengine/sector.cpp -Wall -fvtable-thunks -DMINGW_VOLATILE
-s -O3 -fomit-frame-p
ointer -I. -I./apps -I./include -I./libs -I./plugins -
I./support -Ilibs/libjpeg -Ilibs/libpng -Ilibs/zlib
In file included from .\include\csengine\covtree.h:25,
from libs/csengine/sector.cpp:37:
.\include\csengine\covmask.h: In method `void
csCovMask::SetState(int, int)':
.\include\csengine\covmask.h:97: unknown escape sequence `\i'
.\include\csengine\covmask.h:97: unknown escape sequence `\c'
.\include\csengine\covmask.h:97: unknown escape sequence `\c'
.\include\csengine\covmask.h: In method `int
csCovMask::GetState(int) const':
.\include\csengine\covmask.h:120: unknown escape sequence `\i'
.\include\csengine\covmask.h:120: unknown escape sequence `\c'
.\include\csengine\covmask.h:120: unknown escape sequence `\c'
.\include\csengine\covmask.h: In method `void
csCovMaskTriage::SetState(int, int, int)':
.\include\csengine\covmask.h:281: unknown escape sequence `\i'
.\include\csengine\covmask.h:281: unknown escape sequence `\c'
.\include\csengine\covmask.h:281: unknown escape sequence `\c'
.\include\csengine\covmask.h: In method `int
csCovMaskTriage::GetState(int) const':
.\include\csengine\covmask.h:305: unknown escape sequence `\i'
.\include\csengine\covmask.h:305: unknown escape sequence `\c'

thanks in advance for your help.

Peace,

Paul G.


Nothing real can be threatened.
Nothing unreal exists.
Jorrit Tyberghein
2000-03-15 07:51:49 UTC
Permalink
Post by Paul Garceau
Hi folks,
Not sure what is causing this. Looks like something having to
do with <eol> stuff. Any ideas?
That's very weird. Did you try converting the file to another <eol> format?
i.e. from CR/LF to LF or vice versa?

Greetings,
Post by Paul Garceau
c++ -c -o out/NT4/INTEL/optimize/sector.obj
libs/csengine/sector.cpp -Wall -fvtable-thunks -DMINGW_VOLATILE
-s -O3 -fomit-frame-p
ointer -I. -I./apps -I./include -I./libs -I./plugins -
I./support -Ilibs/libjpeg -Ilibs/libpng -Ilibs/zlib
In file included from .\include\csengine\covtree.h:25,
.\include\csengine\covmask.h: In method `void
.\include\csengine\covmask.h:97: unknown escape sequence `\i'
.\include\csengine\covmask.h:97: unknown escape sequence `\c'
.\include\csengine\covmask.h:97: unknown escape sequence `\c'
.\include\csengine\covmask.h: In method `int
.\include\csengine\covmask.h:120: unknown escape sequence `\i'
.\include\csengine\covmask.h:120: unknown escape sequence `\c'
.\include\csengine\covmask.h:120: unknown escape sequence `\c'
.\include\csengine\covmask.h: In method `void
.\include\csengine\covmask.h:281: unknown escape sequence `\i'
.\include\csengine\covmask.h:281: unknown escape sequence `\c'
.\include\csengine\covmask.h:281: unknown escape sequence `\c'
.\include\csengine\covmask.h: In method `int
.\include\csengine\covmask.h:305: unknown escape sequence `\i'
.\include\csengine\covmask.h:305: unknown escape sequence `\c'
thanks in advance for your help.
Peace,
Paul G.
Nothing real can be threatened.
Nothing unreal exists.
_______________________________________________
Crystal-main mailing list
http://lists.sourceforge.net/mailman/listinfo/crystal-main
--
==============================================================================
***@uz.kuleuven.ac.be, University Hospitals KU Leuven BELGIUM

Voodoo is a very interesting religion for the whole family, even those
members of it who are dead.
-- (Terry Pratchett & Neil Gaiman, Good Omens)
==============================================================================
Paul Garceau
2000-03-16 03:20:49 UTC
Permalink
Hi folks,
Post by Jorrit Tyberghein
Post by Paul Garceau
Hi folks,
Not sure what is causing this. Looks like something
having to
do with <eol> stuff. Any ideas?
That's very weird. Did you try converting the file to another
<eol> format? i.e. from CR/LF to LF or vice versa?
Yes. Same thing happened with igraph2d, except in that case it
was simply "unknown escape sequence '\i'". It occurred at line
117 of igraph2d.

Initial case for situation below was after I'd pulled the
latest update for covmask.h from the repository (update).

Peace,

Paul G.
Post by Jorrit Tyberghein
Greetings,
### Original message follows ###
Post by Jorrit Tyberghein
Post by Paul Garceau
c++ -c -o out/NT4/INTEL/optimize/sector.obj
libs/csengine/sector.cpp -Wall -fvtable-thunks
-DMINGW_VOLATILE -s -O3 -fomit-frame-p ointer -I. -I./apps
-I./include -I./libs -I./plugins - I./support -Ilibs/libjpeg
-Ilibs/libpng -Ilibs/zlib In file included from
.\include\csengine\covtree.h:25,
.\include\csengine\covmask.h: In method `void
.\include\csengine\covmask.h:97: unknown escape sequence `\i'
.\include\csengine\covmask.h:97: unknown escape sequence `\c'
.\include\csengine\covmask.h:97: unknown escape sequence `\c'
.\include\csengine\covmask.h: In method `int
.\include\csengine\covmask.h:120: unknown escape sequence `\i'
.\include\csengine\covmask.h:120: unknown escape sequence `\c'
.\include\csengine\covmask.h:120: unknown escape sequence `\c'
.\include\csengine\covmask.h: In method `void
.\include\csengine\covmask.h:281: unknown escape sequence `\i'
.\include\csengine\covmask.h:281: unknown escape sequence `\c'
.\include\csengine\covmask.h:281: unknown escape sequence `\c'
.\include\csengine\covmask.h: In method `int
.\include\csengine\covmask.h:305: unknown escape sequence `\i'
.\include\csengine\covmask.h:305: unknown escape sequence `\c'
thanks in advance for your help.
Peace,
Paul G.
Nothing real can be threatened.
Nothing unreal exists.
_______________________________________________
Crystal-main mailing list
http://lists.sourceforge.net/mailman/listinfo/crystal-main
--
=================================================================
Hospitals KU Leuven BELGIUM
Voodoo is a very interesting religion for the whole family, even
those members of it who are dead.
-- (Terry Pratchett & Neil Gaiman, Good Omens)
=================================================================
=============
_______________________________________________
Crystal-main mailing list
http://lists.sourceforge.net/mailman/listinfo/crystal-main
Nothing real can be threatened.
Nothing unreal exists.
Paul Garceau
2000-03-20 02:33:39 UTC
Permalink
Hi folks,
Post by Paul Garceau
Not sure what is causing this. Looks like something
having to
do with <eol> stuff.
Turned out to be that CS_CM_8x8 was not defined and CS_CM_4x4
was. Not sure why this would cause a problem, but it did. At
any rate, by commenting out #define CS_CM_4x4, and uncommenting
CS_CM_8x8, the "unknown escape sequence" error disappeared. It
had nothing to do with <eol> stuff.

Now, before I commit the modified file, can someone tell my why
there is a difference?
The masks are obviously different.
What is the purpose for CS_CM_8x8 or CS_CM_4x4? Could find no
reference to either in any of the documentation. It is clear
from the source that CS_CM_xxx governs something called a
coverage mask. This looks like something to do with "culling".

Peace,

Paul G.


Nothing real can be threatened.
Nothing unreal exists.
Jorrit Tyberghein
2000-03-20 07:15:09 UTC
Permalink
Post by Paul Garceau
Hi folks,
Post by Paul Garceau
Not sure what is causing this. Looks like something
having to
do with <eol> stuff.
Turned out to be that CS_CM_8x8 was not defined and CS_CM_4x4
was. Not sure why this would cause a problem, but it did. At
any rate, by commenting out #define CS_CM_4x4, and uncommenting
CS_CM_8x8, the "unknown escape sequence" error disappeared. It
had nothing to do with <eol> stuff.
Now, before I commit the modified file, can someone tell my why
there is a difference?
The masks are obviously different.
What is the purpose for CS_CM_8x8 or CS_CM_4x4? Could find no
reference to either in any of the documentation. It is clear
from the source that CS_CM_xxx governs something called a
coverage mask. This looks like something to do with "culling".
Yes. I would be very very glad if you found a solution which did not involve
removing the #ifdef. Did you try the following:
Instead of:

#if defined (CS_CM_4x4)
#elif defined (CS_CM_8x8)
#endif

Just do:

#ifdef CS_CM_4x4
#else
#endif

Maybe that will also work?

Greetings,


--
==============================================================================
***@uz.kuleuven.ac.be, University Hospitals KU Leuven BELGIUM

"Woof. In tones of low menace."
-- Gaspode the wonder dog
(Terry Pratchett, Moving Pictures)
==============================================================================
Paul Garceau
2000-03-21 00:47:47 UTC
Permalink
Hi folks,

On 20 Mar 00, at 8:15, the Illustrious Jorrit Tyberghein wrote:

(skip)
Post by Jorrit Tyberghein
Post by Paul Garceau
Turned out to be that CS_CM_8x8 was not defined and
CS_CM_4x4
was. Not sure why this would cause a problem, but it did. At
any rate, by commenting out #define CS_CM_4x4, and uncommenting
CS_CM_8x8, the "unknown escape sequence" error disappeared. It
had nothing to do with <eol> stuff.
Now, before I commit the modified file, can someone
tell my why
there is a difference?
The masks are obviously different.
What is the purpose for CS_CM_8x8 or CS_CM_4x4? Could
find no
reference to either in any of the documentation. It is clear
from the source that CS_CM_xxx governs something called a
coverage mask. This looks like something to do with "culling".
Yes. I would be very very glad if you found a solution which did
#if defined (CS_CM_4x4)
#elif defined (CS_CM_8x8)
#endif
Currently I've defined it as follows; primarily due to the fact
that no one else has complained about the problem in their
builds:

Originally (hard-coded, only found in covmask.h):

// Define one of the three below to set the coverage mask size.
#define CS_CM_4x4
//#define CS_CM_8x8

<no "third one" to be found>


My modification/hack (unsubmitted):

// Define one of the three below to set the coverage mask size.
#if defined (__MINGW32__)
//# define CS_CM_4x4
# define CS_CM_8x8
#else
//# define CS_CM_8x8
# define CS_CM_4x4
#endif

Which, in effect reflects the original hard-coding. IF the gcc
is EGCS, djgpp, or Cygwin, the else route is taken, thus
maintaining original hard-code (#define CS_CM_4x4) for anything
that is not Mingw.

Biggest problem I have with this is that Mingw is being used by
others (with win9x) who have not complained or commented about
this problem. For them, they will have to modify, by hand, the
first section of the "#if defined" iff they require the
CS_CM_4x4 #define.
Post by Jorrit Tyberghein
#ifdef CS_CM_4x4
#else
#endif
Maybe that will also work?
Well, I think it would if CS_CM_4x4 was pre-defined somewhere
(at the make level perhaps?), but it isn't. Jorrit, are you
saying that CS_CM_4x4 is supposed to be pre-defined?
NT file searches (by text occurrence, case sensitive) yield no
other file within the CS directory structure which has
"CS_CM_4x4" defined.

It sounds like some of the build environments default to
CS_CM_4x4 (no idea how since first definition for CS_CM_4x4
<#define CS_CM_4x4> only exists in
"include\csengine\covmask.h"), and other build environments do
not.

In libs\csengine\culler\covtree.cpp, there is an #if defined
(CS_CM_8x8). However, compiler doesn't get to covtree.cpp (line
128) until after covmask.h is included (line 21, covtree.cpp).
Post by Jorrit Tyberghein
Greetings,
Peace,

Paul G.


Nothing real can be threatened.
Nothing unreal exists.
Jorrit Tyberghein
2000-03-21 07:26:45 UTC
Permalink
Post by Paul Garceau
Well, I think it would if CS_CM_4x4 was pre-defined somewhere
(at the make level perhaps?), but it isn't. Jorrit, are you
saying that CS_CM_4x4 is supposed to be pre-defined?
Look closer Paul. It is defined in that header. You even pasted the define
into your mail. It is the
#define CS_CM_4x4
at the beginning of include/csengine/covmask.h

Greetings,


--
==============================================================================
***@uz.kuleuven.ac.be, University Hospitals KU Leuven BELGIUM

"Woof. In tones of low menace."
-- Gaspode the wonder dog
(Terry Pratchett, Moving Pictures)
==============================================================================
Paul Garceau
2000-03-21 08:52:55 UTC
Permalink
Post by Jorrit Tyberghein
Post by Paul Garceau
Well, I think it would if CS_CM_4x4 was pre-defined
somewhere
(at the make level perhaps?), but it isn't. Jorrit, are you
saying that CS_CM_4x4 is supposed to be pre-defined?
Look closer Paul. It is defined in that header.
Yes. Looks like a miscommunication here --

CS_CM_4x4 is not defined anywhere outside of covmask.h. Are
you suggesting that CS_CM_4x4 is the default coverage mask
#define for CS? In other words; Is it assumed that all
systems/platforms shall use CS_CM_4x4? If so, then why even
have CS_CM_8x8?

Specifically:

#define CS_CM_4x4
...
#ifndef CS_CM_4x4
# define CS_CM_8x8 <-- line will never be executed
#endif
...

Or...is there something else I'm missing here?

Peace,

Paul G.


Nothing real can be threatened.
Nothing unreal exists.
Jorrit Tyberghein
2000-03-21 08:57:13 UTC
Permalink
Post by Paul Garceau
Yes. Looks like a miscommunication here --
CS_CM_4x4 is not defined anywhere outside of covmask.h. Are
you suggesting that CS_CM_4x4 is the default coverage mask
#define for CS? In other words; Is it assumed that all
systems/platforms shall use CS_CM_4x4? If so, then why even
have CS_CM_8x8?
It is just another way to do the algorithm. Currently I choose to use
CS_CM_4x4 but I have not decided if I'm going to stick with that.
It has nothing to do with platform dependencies or anything. It is just
a variant of the algorithm to select.

Greetings,

--
==============================================================================
***@uz.kuleuven.ac.be, University Hospitals KU Leuven BELGIUM

"Woof. In tones of low menace."
-- Gaspode the wonder dog
(Terry Pratchett, Moving Pictures)
==============================================================================
Paul Garceau
2000-03-21 10:54:24 UTC
Permalink
Post by Jorrit Tyberghein
Post by Paul Garceau
Yes. Looks like a miscommunication here --
CS_CM_4x4 is not defined anywhere outside of covmask.h.
Are
you suggesting that CS_CM_4x4 is the default coverage mask
#define for CS? In other words; Is it assumed that all
systems/platforms shall use CS_CM_4x4? If so, then why even
have CS_CM_8x8?
It is just another way to do the algorithm. Currently I choose to
use CS_CM_4x4 but I have not decided if I'm going to stick with
that. It has nothing to do with platform dependencies or
anything. It is just a variant of the algorithm to select.
Thanks for the clarification, Jorrit. It is appreciated.

Now that we've got that straightened out, I have the current
version of covmask.h set to assume CS_CM_4x4 unless it is hard-
coded otherwise. Exception, if the compiler is Mingw, then CS
assumes CS_CM_8x8 in order to avoid the "unknown escape
sequence" problem.

Peace,

Paul G.
Post by Jorrit Tyberghein
Greetings,
--
=================================================================
Hospitals KU Leuven BELGIUM
"Woof. In tones of low menace."
-- Gaspode the wonder dog
(Terry Pratchett, Moving Pictures)
=================================================================
=============
_______________________________________________
Crystal-main mailing list
http://lists.sourceforge.net/mailman/listinfo/crystal-main
Nothing real can be threatened.
Nothing unreal exists.
Jorrit Tyberghein
2000-03-21 10:59:11 UTC
Permalink
Post by Paul Garceau
Thanks for the clarification, Jorrit. It is appreciated.
Now that we've got that straightened out, I have the current
version of covmask.h set to assume CS_CM_4x4 unless it is hard-
coded otherwise. Exception, if the compiler is Mingw, then CS
assumes CS_CM_8x8 in order to avoid the "unknown escape
sequence" problem.
This is really a terrible solution. I don't understand why you are
having such problems with something that is used ALL over CS.
The construct #if defined(xxx) #else defined (xxx) is used almost
everywhere.

It is really a pity that you have to go to such lengths (even by chosing
another UNTESTED variant of the algorithm). Note that selecting
CS_CM_8x8 may cause bugs in CS. This variant has NEVER been
tested good. It was just included for future research into the possibility
of switching to that mode.

Doesn't the following solution work:

Instead of

# if defined (CS_CM_4x4)
# elif defined (CS_CM_8x8)
#endif

Use:

#ifdef CS_CM_4x4
#else
#endif

That's much cleaner. I would very much appreciate it if you try
that first and see if that solves your problems. In that case you don't
need to do anything special for MingW at all.

Greetings,

--
==============================================================================
***@uz.kuleuven.ac.be, University Hospitals KU Leuven BELGIUM

"Woof. In tones of low menace."
-- Gaspode the wonder dog
(Terry Pratchett, Moving Pictures)
==============================================================================
Paul Garceau
2000-03-21 12:14:47 UTC
Permalink
Post by Jorrit Tyberghein
Post by Paul Garceau
Thanks for the clarification, Jorrit. It is
appreciated.
Now that we've got that straightened out, I have the
current
version of covmask.h set to assume CS_CM_4x4 unless it is hard-
coded otherwise. Exception, if the compiler is Mingw, then CS
assumes CS_CM_8x8 in order to avoid the "unknown escape
sequence" problem.
This is really a terrible solution. I don't understand why you
are having such problems with something that is used ALL over CS.
The construct #if defined(xxx) #else defined (xxx) is used almost
everywhere.
It is really a pity that you have to go to such lengths (even by
chosing another UNTESTED variant of the algorithm). Note that
selecting CS_CM_8x8 may cause bugs in CS. This variant has NEVER
been tested good. It was just included for future research into
the possibility of switching to that mode.
Instead of
# if defined (CS_CM_4x4)
# elif defined (CS_CM_8x8)
#endif
#ifdef CS_CM_4x4
#else
#endif
That's much cleaner. I would very much appreciate it if you try
that first and see if that solves your problems. In that case you
don't need to do anything special for MingW at all.
Will give it a try...

Peace,

Paul G.
Post by Jorrit Tyberghein
Greetings,
--
=================================================================
Hospitals KU Leuven BELGIUM
"Woof. In tones of low menace."
-- Gaspode the wonder dog
(Terry Pratchett, Moving Pictures)
=================================================================
=============
_______________________________________________
Crystal-main mailing list
http://lists.sourceforge.net/mailman/listinfo/crystal-main
Nothing real can be threatened.
Nothing unreal exists.
Jason Platt
2000-03-21 20:26:57 UTC
Permalink
Ok, I've got a question or two:

What is CS_CM_4x4 and CS_CM_8x8 used for? (the technical explination
please)


=====
Jason Platt.

"In theory: theory and practice are the same.
In practice: they arn't."

ICQ# 1546328

__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com
Jorrit Tyberghein
2000-03-22 07:18:48 UTC
Permalink
Post by Jason Platt
What is CS_CM_4x4 and CS_CM_8x8 used for? (the technical explination
please)
It is for the coverage mask tree. The 4x4 and 8x8 defines allow you to select
the mask resolution of that tree. Default is 4x4 which means that at every level
of the tree you divide in 4x4 areas and thus have 16 children. With 8x8 you would
have 8x8 children.

Greetings,


--
==============================================================================
***@uz.kuleuven.ac.be, University Hospitals KU Leuven BELGIUM

"When a man is tired of Ankh-Morpork, he is tired of ankle-deep slurry."
-- (Terry Pratchett, Mort)
==============================================================================
Jason Platt
2000-03-22 18:07:00 UTC
Permalink
Post by Jason Platt
Post by Jason Platt
What is CS_CM_4x4 and CS_CM_8x8 used for? (the technical
explination
Post by Jason Platt
please)
It is for the coverage mask tree. The 4x4 and 8x8 defines allow you
to select
the mask resolution of that tree. Default is 4x4 which means that at
every level
of the tree you divide in 4x4 areas and thus have 16 children. With
8x8 you would
have 8x8 children.
Greetings,
Ahh, that helps! Thank you.

=====
Jason Platt.

"In theory: theory and practice are the same.
In practice: they arn't."

ICQ# 1546328

__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com

Continue reading on narkive:
Loading...