Re: LyX + MinGW + MinSYS + GPLed QT/Win32 --- success!

[Lars Gullik Bjønnes]

Angus Leeming <[EMAIL PROTECTED]> writes:

| I spent a good chunk of this weekend trying to get LyX to compile on
| Windows using the Qt port from the cygwin-kde people and I've finally
| succeeded. Attached are some notes detailing what I needed to do, together
| with a patch and os_win32.h which should go in src/support. The patch is
| essentially Ruurd's, but I've put it together piece by piece as
| compilation failed.
>
| The resulting executable is rather neutered because the fork/execvp stuff
| is extremely ropey. It compiles, but the whole waitpid thing is guaranteed
| not to work ;-)
>
| What I'd like to do is to break this patch into smaller parts and apply
| each piece with a proper explanation about what it does and why its
| needed. I've only compiled lyx-13x, but much of the patch will apply
| without changes to the 14x tree too.
>
| So, JMarc and Lars. Are you happy if I put together small patches for the
| 13x and 14x trees and proceed SLOWLY ?

Sure... but slowly.  We should make sure that the resulting code is
as-clean or cleaner than what we have now before comitting.

-- 
Lgb

Re: LyX + MinGW + MinSYS + GPLed QT/Win32 --- success!

[Juergen Spitzmueller]

Angus Leeming wrote:

> I spent a good chunk of this weekend trying to get LyX to compile on
> Windows using the Qt port from the cygwin-kde people and I've finally
> succeeded.

cool!

Jürgen

LyX + MinGW + MinSYS + GPLed QT/Win32 --- success!

[Angus Leeming]

I spent a good chunk of this weekend trying to get LyX to compile on
Windows using the Qt port from the cygwin-kde people and I've finally
succeeded. Attached are some notes detailing what I needed to do, together
with a patch and os_win32.h which should go in src/support. The patch is
essentially Ruurd's, but I've put it together piece by piece as
compilation failed.

The resulting executable is rather neutered because the fork/execvp stuff
is extremely ropey. It compiles, but the whole waitpid thing is guaranteed
not to work ;-)

What I'd like to do is to break this patch into smaller parts and apply
each piece with a proper explanation about what it does and why its
needed. I've only compiled lyx-13x, but much of the patch will apply
without changes to the 14x tree too.

So, JMarc and Lars. Are you happy if I put together small patches for the
13x and 14x trees and proceed SLOWLY ?

-- 
Angus? src/support/os_win32.h

Index: src/BufferView_pimpl.C

===

RCS file: /cvs/lyx/lyx-devel/src/BufferView_pimpl.C,v

retrieving revision 1.322.2.3

diff -u -r1.322.2.3 BufferView_pimpl.C

--- src/BufferView_pimpl.C	2004/12/07 10:48:01	1.322.2.3

+++ src/BufferView_pimpl.C	2004/12/13 19:46:24

@@ -63,7 +63,9 @@

 #include "BoostFormat.h"

 

 #include 

+#ifndef __MINGW32__

 #include 

+#endif

 

 

 using std::vector;

Index: src/LaTeX.C

===

RCS file: /cvs/lyx/lyx-devel/src/LaTeX.C,v

retrieving revision 1.74.2.3

diff -u -r1.74.2.3 LaTeX.C

--- src/LaTeX.C	2004/12/07 10:48:11	1.74.2.3

+++ src/LaTeX.C	2004/12/13 22:09:26

@@ -380,11 +380,7 @@

 

 int LaTeX::operator()()

 {

-#ifndef __EMX__

-	string tmp = cmd + ' ' + QuoteName(file) + " > /dev/null";

-#else // cmd.exe (OS/2) causes SYS0003 error at "/dev/null"

-	string tmp = cmd + ' ' + file + " > nul";

-#endif

+	string tmp = cmd + ' ' + QuoteName(file) + " > " + os::nulldev_;

 	Systemcall one;

 	return one.startscript(Systemcall::Wait, tmp);

 }

Index: src/Makefile.am

===

RCS file: /cvs/lyx/lyx-devel/src/Makefile.am,v

retrieving revision 1.151

diff -u -r1.151 Makefile.am

--- src/Makefile.am	2003/01/12 22:20:21	1.151

+++ src/Makefile.am	2004/12/13 17:48:30

@@ -219,14 +219,14 @@

 	vspace.C \

 	vspace.h

 

-lyx_main.o: lyx_main.C lyx_main.h config.h version.h \

+lyx_main.$(OBJEXT): lyx_main.C lyx_main.h config.h version.h \

 	lyxrc.h support/path.h support/filetools.h \

 	bufferlist.h debug.h support/FileInfo.h lastfiles.h intl.h \

 	lyxserver.h layout.h gettext.h kbmap.h commandtags.h language.h

 	$(CXXCOMPILE) -DLYX_DIR=\"$(pkgdatadir)\" \

 	  -DTOP_SRCDIR=\"$(top_srcdir)\" -c $(top_srcdir)/src/lyx_main.C

 

-main.o: main.C config.h lyx_main.h gettext.h LString.h support/filetools.h support/os.h

+main.$(OBJEXT): main.C config.h lyx_main.h gettext.h LString.h support/filetools.h support/os.h

 	$(CXXCOMPILE) -DLOCALEDIR=\"$(localedir)\" -c $(top_srcdir)/src/main.C

 

 dist-hook:

Index: src/ispell.C

===

RCS file: /cvs/lyx/lyx-devel/src/ispell.C,v

retrieving revision 1.5.2.5

diff -u -r1.5.2.5 ispell.C

--- src/ispell.C	2004/12/07 10:48:23	1.5.2.5

+++ src/ispell.C	2004/12/13 20:23:16

@@ -26,6 +26,7 @@

 #include 

 #endif

 #include 

+#include 

 

 #ifndef CXX_GLOBAL_CSTD

 using std::strcpy;

@@ -73,6 +74,9 @@

 

 int LaunchIspell::generateChild()

 {

+#ifdef _WIN32

+	return -1;

+#else

 	pid_t isp_pid = fork();

 

 	if (isp_pid != 0) {

@@ -173,6 +177,7 @@

 

 	lyxerr << "LyX: Failed to start ispell!" << endl;

 	_exit(0);

+#endif

 }

 

 

@@ -182,6 +187,7 @@

 ISpell::ISpell(BufferParams const & params, string const & lang)

 	: in(0), out(0), inerr(0), str(0)

 {

+#ifndef _WIN32

 	lyxerr[Debug::GUI] << "Created ispell" << endl;

 

 	// static due to the setvbuf. Ugly.

@@ -268,6 +274,7 @@

 	close(pipeerr[1]);

 	child_->kill();

 	child_.reset(0);

+#endif

 }

 

 

@@ -300,6 +307,9 @@

 

 bool ISpell::select(bool & err_read)

 {

+#ifdef _WIN32

+	return false;

+#else

 	fd_set infds;

 	struct timeval tv;

 	int retval = 0;

@@ -328,6 +338,7 @@

 	fgets(buf, BUFSIZ, in);

 	err_read = false;

 	return false;

+#endif

 }

 

 

Index: src/lyx_main.C

===

RCS file: /cvs/lyx/lyx-devel/src/lyx_main.C,v

retrieving revision 1.134.2.4

diff -u -r1.134.2.4 lyx_main.C

--- src/lyx_main.C	2004/12/07 10:48:26	1.134.2.4

+++ src/lyx_main.C	2004/12/13 23:21:30

@@ -289,7 +289,7 @@

 	do {

 		// Path of binary/../share/name of binary/

 		searchpath += NormalizePath(AddPath(binpath, "../share/") +

-		  OnlyFilename(binname)) + ';';

+		  ChangeExtension(OnlyFilename(binname),"")) + ';';

 

 		// Follow Symlinks

 		FileInfo file(fullbinpath, true);

@@ -434,7 +434,7 @@

 

 		// default behav

Re: The first configure test...

[Angus Leeming]

Lars Gullik Bjønnes wrote:
> who is lyx::mkdir used. Do we every use any fancy mode values?

Well, we do use different mode values:

src/support/filetools.C:
 if (tmpfl.empty() || lyx::mkdir(tmpfl, 0700)) {
 if (lyx::mkdir(tmpfl, 0777)) {
 if (lyx::mkdir(deflt, 0777)) {
bool createDirectory(string const & path, int permission) {
 ...
 if (lyx::mkdir(temp, permission)) {

src/lyx_main.C: if (!createDirectory(user_lyxdir, 0755)) {

So, I guess that the interface should stay as-is.

> | #include 
> | #include 
>>
> | int main() {
> | int const success = mkdir("foo", 0777);
> | return 0;
> | }
>>
> | Is there an autotool-standard way to do this?
> 
> Hmm... kindo I guess. Jean-Marc might have the answer close by :-)

-- 
Angus

Re: [PATCHES 13x, 14x] Typo in os.C?

[Angus Leeming]

Jean-Marc Lasgouttes wrote:
> Angus> I think you'll find that Ruurd's patch has some more #ifdef
> Angus> magic in os_win32.C that differentiates between native Windows
> Angus> and Cygwin. At the moment, os_win32 really means cygwin. I
> Angus> guess that we could have os_cygwin.C and os_win32.C if that
> Angus> makes life more understandable.
> 
> I think I would prefer that, if it is easily doable.

OK.

-- 
Angus

The first configure test...

[Angus Leeming]

Consider this patch:

 int lyx::mkdir(string const & pathname, unsigned long int mode)
 {
  // FIXME: why don't we have mode_t in lyx::mkdir prototype ??
+#ifdef _WIN32
+ return ::mkdir(pathname.c_str());
+#else
  return ::mkdir(pathname.c_str(), mode_t(mode));
+#endif
 }

I guess that configure should generate something like this for a macro
MKDIR_HAS_TWO_ARGUMENTS

#include 
#include 

int main() {
int const success = mkdir("foo", 0777);
return 0;
}

Is there an autotool-standard way to do this?

-- 
Angus

Re: [PATCHES 13x, 14x] Typo in os.C?

[Jean-Marc Lasgouttes]

> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:

Angus> I think you'll find that Ruurd's patch has some more #ifdef
Angus> magic in os_win32.C that differentiates between native Windows
Angus> and Cygwin. At the moment, os_win32 really means cygwin. I
Angus> guess that we could have os_cygwin.C and os_win32.C if that
Angus> makes life more understandable.

I think I would prefer that, if it is easily doable.

JMarc

Re: [PATCH 13x, 14x]: Typo in systemcall.C?

[Jean-Marc Lasgouttes]

> "Lars" == Lars Gullik Bjønnes <[EMAIL PROTECTED]> writes:

Lars> Is the OS/2 port completely dead now?

That's a very good question. We may want to ask SMiyata, but my
googling did not find anything newer than 1.1.6fix3.

JMarc

Re: Binary mode ifstream.

[Lars Gullik Bjønnes]

Angus Leeming <[EMAIL PROTECTED]> writes:

| Ruurd has the following one line patch:
>
| Index: src/support/copy.C
|  
|  bool lyx::copy(string const & from, string const & to)
|  {
| - ifstream ifs(from.c_str());
| + ifstream ifs(from.c_str(), ios::binary);
|   if (!ifs)
|return false;
|   ofstream ofs(to.c_str(),
>
| Googling comes up with this:
>
| http://gcc.gnu.org/ml/gcc/1998-11/msg00258.html
| Under unix, binary file IO is always the case.  Note that in the 
| iostreams library, binary means the same thing it means in ftp ... do 
| not convert line ends.  therefore:
>
| Sinse we don't support network protocols like ftp, I think that this 
| patch is entirely safe. Agreed?

Yes. and even with ftp it would be fine.

-- 
Lgb

Re: [PATCH 13x, 14x]: Typo in systemcall.C?

[Lars Gullik Bjønnes]

Angus Leeming <[EMAIL PROTECTED]> writes:

| Ruurd Reitsma wrote:
>>> Ruurd has changed the invocation of system() under DOS to:
>>>  system("start /min ");
>>> from
>>>  system("start /min/n ");
>>>
>>> Ruurd, what's the reason for this? Is it just a typo in the original
>>> code?
>> 
>> The /n came from the OS/2 port. Makes no sense on Windows...
>
| Fine. Then I'll commit it.

Is the OS/2 port completely dead now?

-- 
Lgb

Re: [PATCH 13x, 14x]: Typo in systemcall.C?

[Angus Leeming]

Ruurd Reitsma wrote:
>> Ruurd has changed the invocation of system() under DOS to:
>>  system("start /min ");
>> from
>>  system("start /min/n ");
>>
>> Ruurd, what's the reason for this? Is it just a typo in the original
>> code?
> 
> The /n came from the OS/2 port. Makes no sense on Windows...

Fine. Then I'll commit it.

Incidentally, should I CC you with these messages or are you reading gmane?

-- 
Angus

Re: [PATCHES 13x, 14x] Typo in os.C?

[Jean-Marc Lasgouttes]

I am not sure that I understand the logic of using os_win32.C for
cygwin and mingw. os_win32.C uses some cygwin functions, so I would
think that it is not good for the mingw version.

Actually, cygwin should be considered to be unix, since it is posix...

What is special about cygwin, except from the fact that it needs a way
to translate file names to/from dos conventions?

As you can see, I am a bit lost already.

JMarc

Re: sideways tables/figures

[Juergen Spitzmueller]

Georg Baum wrote:

> Jürgen, I thought you fixed that in the meantime?

No, not yet. I haven't cvs access right now. Can you do it? You just have to
remove the #if 0 conditional.

Thanks,
Jürgen

Re: [PATCH 13x, 14x]: Typo in systemcall.C?

[Ruurd Reitsma]

"Angus Leeming" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> Ruurd has changed the invocation of system() under DOS to:
>  system("start /min ");
> from
>  system("start /min/n ");
>
> Ruurd, what's the reason for this? Is it just a typo in the original
> code?

The /n came from the OS/2 port. Makes no sense on Windows...

Starts a separate window to run a specified program or command.

START ["title"] [/Dpath] [/I] [/MIN] [/MAX] [/SEPARATE | /SHARED]
  [/LOW | /NORMAL | /HIGH | /REALTIME | /ABOVENORMAL | /BELOWNORMAL]
  [/WAIT] [/B] [command/program]
  [parameters]

"title" Title to display in  window title bar.
pathStarting directory
B   Start application without creating a new window. The
application has ^C handling ignored. Unless the application
enables ^C processing, ^Break is the only way to interrupt
the application
I   The new environment will be the original environment passed
to the cmd.exe and not the current environment.
MIN Start window minimized
MAX Start window maximized
SEPARATEStart 16-bit Windows program in separate memory space
SHARED  Start 16-bit Windows program in shared memory space
LOW Start application in the IDLE priority class
NORMAL  Start application in the NORMAL priority class
HIGHStart application in the HIGH priority class
REALTIMEStart application in the REALTIME priority class
ABOVENORMAL Start application in the ABOVENORMAL priority class
BELOWNORMAL Start application in the BELOWNORMAL priority class
WAITStart application and wait for it to terminate
command/program

[PATCH 13x, 14x]: Typo in systemcall.C?

[Angus Leeming]

Ruurd has changed the invocation of system() under DOS to:
 system("start /min ");
from
 system("start /min/n ");

Ruurd, what's the reason for this? Is it just a typo in the original 
code?

Angus
Index: src/support/ChangeLog
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/support/ChangeLog,v
retrieving revision 1.149.2.11
diff -u -p -r1.149.2.11 ChangeLog
--- src/support/ChangeLog	14 Dec 2004 10:40:09 -	1.149.2.11
+++ src/support/ChangeLog	14 Dec 2004 12:10:13 -
@@ -1,5 +1,12 @@
 2004-12-14  Angus Leeming  <[EMAIL PROTECTED]>
 
+	* os.C: Add _WIN32 to the #define.
+ 
+	* systemcall.C (startscript): remove trailing '/n' from request to
+	start the command in a minimized window under DOS.
+
+2004-12-14  Angus Leeming  <[EMAIL PROTECTED]>
+
 	* Makefile.am (INCLUDES): Remove trailing slash from -Ifoo/
 	to overcome bug in the MinGW port of g++ 3.4.2.
 
Index: src/support/os.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/support/os.C,v
retrieving revision 1.3.2.1
diff -u -p -r1.3.2.1 os.C
--- src/support/os.C	7 Dec 2004 10:50:09 -	1.3.2.1
+++ src/support/os.C	14 Dec 2004 12:10:13 -
@@ -2,7 +2,7 @@
 
 #ifdef __EMX__
 #include "os_os2.C"
-#elif defined(__CYGWIN__) || defined(__CYGWIN32__)
+#elif defined(__CYGWIN__) || defined(__CYGWIN32__) || defined(_WIN32)
 #include "os_win32.C"
 #else
 #include "os_unix.C"
Index: src/support/systemcall.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/support/systemcall.C,v
retrieving revision 1.4.2.1
diff -u -p -r1.4.2.1 systemcall.C
--- src/support/systemcall.C	7 Dec 2004 10:50:09 -	1.4.2.1
+++ src/support/systemcall.C	14 Dec 2004 12:10:13 -
@@ -28,10 +28,13 @@ int Systemcall::startscript(Starttype ho
 	string command = what;
 
 	if (how == DontWait) {
-		if (os::shell() == os::UNIX) {
+		switch (os::shell()) {
+		case os::UNIX:
 			command += " &";
-		} else {
-			command = "start /min/n " + command;
+			break;
+		case os::CMD_EXE:
+			command = "start /min " + command;
+			break;
 		}
 	}
 
? src/support/lyxlib-orig.C
? src/support/lyxlib.C
Index: src/support/ChangeLog
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/support/ChangeLog,v
retrieving revision 1.280
diff -u -p -r1.280 ChangeLog
--- src/support/ChangeLog	14 Dec 2004 10:41:08 -	1.280
+++ src/support/ChangeLog	14 Dec 2004 12:10:45 -
@@ -1,5 +1,12 @@
 2004-12-14  Angus Leeming  <[EMAIL PROTECTED]>
 
+	* os.C: Add _WIN32 to the #define.
+
+	* systemcall.C (startscript): remove trailing '/n' from request to
+	start the command in a minimized window under DOS.
+
+2004-12-14  Angus Leeming  <[EMAIL PROTECTED]>
+
 	* Makefile.am (AM_CPPFLAGS): Remove trailing slash from -Ifoo/
 	to overcome bug in the MinGW port of g++ 3.4.2.
 
Index: src/support/os.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/support/os.C,v
retrieving revision 1.6
diff -u -p -r1.6 os.C
--- src/support/os.C	23 Aug 2003 00:16:57 -	1.6
+++ src/support/os.C	14 Dec 2004 12:10:45 -
@@ -12,7 +12,7 @@
 
 #ifdef __EMX__
 #include "os_os2.C"
-#elif defined(__CYGWIN__) || defined(__CYGWIN32__)
+#elif defined(__CYGWIN__) || defined(__CYGWIN32__) || defined(_WIN32)
 #include "os_win32.C"
 #else
 #include "os_unix.C"
Index: src/support/systemcall.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/support/systemcall.C,v
retrieving revision 1.9
diff -u -p -r1.9 systemcall.C
--- src/support/systemcall.C	7 Nov 2004 13:22:51 -	1.9
+++ src/support/systemcall.C	14 Dec 2004 12:10:45 -
@@ -33,10 +33,13 @@ int Systemcall::startscript(Starttype ho
 	string command = what;
 
 	if (how == DontWait) {
-		if (os::shell() == os::UNIX) {
+		switch (os::shell()) {
+		case os::UNIX:
 			command += " &";
-		} else {
-			command = "start /min/n " + command;
+			break;
+		case os::CMD_EXE:
+			command = "start /min " + command;
+			break;
 		}
 	}
 

Binary mode ifstream.

[Angus Leeming]

Ruurd has the following one line patch:

Index: src/support/copy.C
 
 bool lyx::copy(string const & from, string const & to)
 {
- ifstream ifs(from.c_str());
+ ifstream ifs(from.c_str(), ios::binary);
  if (!ifs)
   return false;
  ofstream ofs(to.c_str(),

Googling comes up with this:

http://gcc.gnu.org/ml/gcc/1998-11/msg00258.html
Under unix, binary file IO is always the case.  Note that in the 
iostreams library, binary means the same thing it means in ftp ... do 
not convert line ends.  therefore:

Sinse we don't support network protocols like ftp, I think that this 
patch is entirely safe. Agreed?

Angus

Re: LyX + MinGW + MinSYS + GPLed QT/Win32 --- success!

[Jean-Marc Lasgouttes]

> "Ruurd" == Ruurd Reitsma <[EMAIL PROTECTED]> writes:

>> > Do the math fonts work correctly? Ruurd once said they were too
>> large.
>> 

Ruurd> They work since the GPLed Qt moved to version 3.3.3. Some are
Ruurd> still not in the right place however.

I do not remember whether I already pointed it out, but there is a
hack in lyx_gui.C for LyX/Mac that fixes such a problem. This is
lyx_gui::needs_ugly_metrics_hack() in src/frontends/qt2/lyx_gui.C.
Did you try to make it return 'true' for LyX/Win?

>> > How come you do not need the dirent stuff?
>> 
>> Shrug. Dunno.
>> >

Ruurd> It´s already included in mingw, not in MSVC.

Ahh, OK. I guess it would be a good idea to incorporate patches for
mingw first, and then msvc. Is that doable?

JMarc

Re: [PATCHES 13x, 14x] Typo in os.C?

[Jean-Marc Lasgouttes]

> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:

Angus> I don't think there's any need to drop support for old cygwin,
Angus> so I propose to add the _WIN32 bit in addition to the original
Angus> call. Ok?

It would be nice to know what _WIN32 really is. All windows compilers?
MinGW? MSVC?

For example, boost has:

#elif defined(_WIN32) || defined(__WIN32__) || defined(WIN32)
// win32:
#  define BOOST_PLATFORM_CONFIG "boost/config/platform/win32.hpp"

JMarc

Re: sideways tables/figures

[Edwin Leuven]

It is safely testable. :-) 
you must be kidding ;-)
Configure it to add the version
number to the program name as a suffix, that way
the "lyx1.4cvs" binary doesn't overwrite your
good old "lyx" binary.
that's what i did
thanks, ed.
PS first problem:
set/clear all borders in tabulars doesn't work...
PPS trying to open a doc, lyx crashes:
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread -1219886976 (LWP 2229)]
LyXText::setCounter (this=0x8656c1c, [EMAIL PROTECTED], pit=66)
at ../../lyx-devel/src/text2.C:793
793 if (in->lyxCode() == 
InsetBase::FLOAT_CODE ||
(gdb) bt
#0  LyXText::setCounter (this=0x8656c1c, [EMAIL PROTECTED], pit=66)
at ../../lyx-devel/src/text2.C:793
#1  0x08179f75 in LyXText::updateCounters (this=0x8656c1c)
at ../../lyx-devel/src/text2.C:867
#2  0x08176972 in LyXText::init (this=0x8656c1c, bv=0x5f)
at ../../lyx-devel/src/text2.C:95
#3  0x08069fe1 in BufferView::Pimpl::setBuffer (this=0x85b92a0, b=0x85b937c)
at ../../lyx-devel/src/BufferView_pimpl.C:352
#4  0x08069bf9 in BufferView::Pimpl::loadLyXFile (this=0x85b92a0, 
[EMAIL PROTECTED],
tolastfiles=true) at ../../lyx-devel/src/BufferView_pimpl.C:289
#5  0x08063d3c in BufferView::loadLyXFile (this=0x0, [EMAIL PROTECTED])
at ../../lyx-devel/src/BufferView.C:120
#6  0x0811b129 in LyXFunc::open (this=0x85b4140, [EMAIL PROTECTED])
at ../../lyx-devel/src/lyxfunc.C:1651
#7  0x081159cb in LyXFunc::dispatch (this=0x85b4140, [EMAIL PROTECTED])
at ../../lyx-devel/src/lyxfunc.C:990
#8  0x082b27a5 in lyx::frontend::QLToolbar::clicked (this=0x85dff40) at 
LyXView.h:83
#9  0x08326062 in lyx::frontend::QLToolbar::qt_invoke (this=0x85dff40, 
_id=2, _o=0x0)
at QLToolbar_moc.C:171
#10 0xb7b6352c in QObject::activate_signal () from /usr/lib/libqt-mt.so.3

Re: [PATCHES 13x, 14x] Typo in os.C?

[Angus Leeming]

Jean-Marc Lasgouttes wrote:
> Angus> This looks like a simple typo.
> 
> This is not really a typo, but some old versions of cygwin used to
> define __CYGWIN32__, now they define __CYGWIN__ instead.
> 
> See here for example:
> http://sources.redhat.com/ml/gdb-patches/2002-06/msg00130.html
> 
> So your patch drops support for old cygwin and adds support for MINGW,
> I think.
> 
> So the patch _may_ be right :)

Thanks, JMarc. Interestingly, these are the only places we use the macro in
the 1.3.x tree:
$ grep -r 'CYGWIN' src
src/support/os.C:#elif defined(__CYGWIN__) || defined(_WIN32)
src/config.h.in:#if defined(__CYGWIN__) || defined(__CYGWIN32__)

In the 1.4.x tree only the os.C match remains.

I don't think there's any need to drop support for old cygwin, so I propose
to add the _WIN32 bit in addition to the original call. Ok?

-- 
Angus

Re: LyX + MinGW + MinSYS + GPLed QT/Win32 --- success!

[Angus Leeming]

Ruurd Reitsma wrote:
>> os_win32.h is a kludge. It is performing the same rôle as config.h. I
> think
>> that these things should go in config.h, which suggests a
>> config/win32.m4 file.
> 
> It´s a kludge indeed. How come you don´t need the extra posix headers I
> have in my patch?

Hi, Ruurd.

I needed the wait.h stuff, but hacked it into os_win32.h. As for the rest,
I dunno. I don't propose to propogate the wait.h stuff into the LyX tree.
This stuff needs rewriting using CreateProcess et al.

-- 
Angus

Re: [PATCHES 13x, 14x] Typo in os.C?

[Jean-Marc Lasgouttes]

> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:

Angus> This looks like a simple typo. 

This is not really a typo, but some old versions of cygwin used to
define __CYGWIN32__, now they define __CYGWIN__ instead.

See here for example:
http://sources.redhat.com/ml/gdb-patches/2002-06/msg00130.html

So your patch drops support for old cygwin and adds support for MINGW,
I think. 

So the patch _may_ be right :)

JMarc

Re: sideways tables/figures

[Helge Hafting]

Edwin Leuven wrote:
Yes. And if you are willing to test the 1.4 version (and compare with 
your stuff), we can improve the support in the 1.5 cycle :-)

can 1.4cvs be safely used for real work now?
ed.
It is safely testable. :-)  Configure it to add the version
number to the program name as a suffix, that way
the "lyx1.4cvs" binary doesn't overwrite your
good old "lyx" binary.
Helge Hafting

Re: sideways tables/figures

[Jean-Marc Lasgouttes]

> "Juergen" == Juergen Spitzmueller <[EMAIL PROTECTED]> writes:

Juergen> Jean-Marc Lasgouttes wrote:
>> Juergen, concerning the sideways feature, do you know about the
>> rotfloat.sty package, which allows to rotate all kind of floats?

Juergen> I am aware that it exists, but I haven't digged into it very
Juergen> much. Is it supposed to be installed on any LaTeX system? 

I'm not sure but it is installed on my canonical old TeX installation
here :)

Juergen> And does it provide the same features as rotating.sty so that
Juergen> a switch would be possible without too much lost?

As far as I know, it uses rotating.sty and float.sty. 

Anyway, I am not saying that we should implement it right now. Just
wanting to point to it before I forget...

JMarc

Re: sideways tables/figures

[Juergen Spitzmueller]

Juergen Spitzmueller wrote:

>> Juergen, concerning the sideways feature, do you know about the
>> rotfloat.sty package, which allows to rotate all kind of floats?

Answering my own questions:

> I am aware that it exists, but I haven't digged into it very much. Is it
> supposed to be installed on any LaTeX system? 

I think it should be (at least it is included in teTeX).

> And does it provide the same 
> features as rotating.sty so that a switch would be possible without too
> much lost?

It seems to be just an extension to rotating.sty (and loads rotating by
itself). In fact, it uses the same commands (sidewaysfigure,
sidewaystable). So I'd propose to stick with rotating for standard floats
and load rotfloat instead for nonstandard floats. Then the switch should
cause no harms. I'll put this on my agenda-1.5.

Jürgen

Re: sideways tables/figures

[Jean-Marc Lasgouttes]

> "Juergen" == Juergen Spitzmueller <[EMAIL PROTECTED]> writes:

Juergen> It seems to be just an extension to rotating.sty (and loads
Juergen> rotating by itself). In fact, it uses the same commands
Juergen> (sidewaysfigure, sidewaystable). So I'd propose to stick with
Juergen> rotating for standard floats and load rotfloat instead for
Juergen> nonstandard floats. Then the switch should cause no harms.
Juergen> I'll put this on my agenda-1.5.

Good.

JMarc

Re: sideways tables/figures

[Juergen Spitzmueller]

Jean-Marc Lasgouttes wrote:

> Juergen, concerning the sideways feature, do you know about the
> rotfloat.sty package, which allows to rotate all kind of floats?

I am aware that it exists, but I haven't digged into it very much. Is it
supposed to be installed on any LaTeX system? And does it provide the same
features as rotating.sty so that a switch would be possible without too
much lost?

Jürgen

Re: [PATCHES 13x, 14x] Work arounds for broken g++ 3.4 on MinGW

[Jean-Marc Lasgouttes]

> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:

Angus> The MinGW g++ 3.4 appears to have a bug related to -I
Angus> directives. An instruction like: g++ -Ifoo/bar/ -c baz.C will
Angus> fail to find any header files in the foo/bar directory. In
Angus> order to succeed, the trailing '/' must be removed. Ie, this
Angus> works: g++ -Ifoo/bar -c baz.C

Angus> I've reported this on the MinGW-MSYS list, but meanwhile the
Angus> patch to the various Makefile.am files is trivial.

That's OK with me. Actually I prefer the version without slash (and
this is what we do when doing -I$(srcdir), for example.

JMarc

Re: Equation crashes cvs-1.4.0 (12/12/04)

[Georg Baum]

Fred, thanks for the report. This will bite more users if it does not get
fixed.

Fred Nastos wrote:

> Hi,
> 
> I downloaded and built the cvs version last night.  Athough I can
> start it up and create new documents, it seems to crash (without
> starting up) on almost all my documents from 1.3.5.  I thought you
> might want to see this, so I managed to pare one document down
> to a single equation as an example, which I attach.  It seems like
> it's the multiline equations that it doesn't like.

No it is much simpler: Math in an mbox does not work anymore. Attached a
simpler test case that uses the following (legal) latex construct:

$\mbox{$v$}$

André, could you please have a look? I see that the differences between 1.3
and 1.4 in math_parser.C (where the error occurs) are small, but since I
don't know why they were introduced I can't tell how to fix this.


Georg

xx3.lyx
Description: application/lyx

[PATCH 13x] .o --> $(OBJEXT) in Makefile.am

[Angus Leeming]

This one is truly trivial. Moreover, it doesn't appear to be necessary in
1.4.x.

Ok to apply?

-- 
AngusIndex: src/ChangeLog
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/ChangeLog,v
retrieving revision 1.1021.2.49
diff -u -p -r1.1021.2.49 ChangeLog
--- src/ChangeLog	7 Dec 2004 10:48:05 -	1.1021.2.49
+++ src/ChangeLog	14 Dec 2004 09:49:44 -
@@ -1,3 +1,9 @@
+2004-12-14  Angus Leeming  <[EMAIL PROTECTED]>
+
+	* Makefile.am: change targets lyx_main.o and main.o to
+	lyx_main.$(OBJEXT) and main.$(OBJEXT) to ensure compilation
+	proceeds on Windows machines.
+
 2004-12-07  Angus Leeming  <[EMAIL PROTECTED]>
 
 	* *.[Ch]: remove all traces of #pragma interface/implementation
Index: src/Makefile.am
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/Makefile.am,v
retrieving revision 1.151
diff -u -p -r1.151 Makefile.am
--- src/Makefile.am	12 Jan 2003 22:20:21 -	1.151
+++ src/Makefile.am	14 Dec 2004 09:49:44 -
@@ -219,14 +219,15 @@ lyx_SOURCES = \
 	vspace.C \
 	vspace.h
 
-lyx_main.o: lyx_main.C lyx_main.h config.h version.h \
+lyx_main.$(OBJEXT): lyx_main.C lyx_main.h config.h version.h \
 	lyxrc.h support/path.h support/filetools.h \
 	bufferlist.h debug.h support/FileInfo.h lastfiles.h intl.h \
 	lyxserver.h layout.h gettext.h kbmap.h commandtags.h language.h
 	$(CXXCOMPILE) -DLYX_DIR=\"$(pkgdatadir)\" \
 	  -DTOP_SRCDIR=\"$(top_srcdir)\" -c $(top_srcdir)/src/lyx_main.C
 
-main.o: main.C config.h lyx_main.h gettext.h LString.h support/filetools.h support/os.h
+main.$(OBJEXT): main.C config.h lyx_main.h gettext.h LString.h \
+	support/filetools.h support/os.h
 	$(CXXCOMPILE) -DLOCALEDIR=\"$(localedir)\" -c $(top_srcdir)/src/main.C
 
 dist-hook:

Re: [LyX/Mac] Setting PATH in property file.

[Bennett Helm]

On Dec 10, 2004, at 9:49 AM, Jean-Marc Lasgouttes wrote:
However, a possibility would be to remove all this special code for
LyX/Mac in the startup code, and replace it with setting user- and
lyx-dir from the script.
I've done this -- or at least tried to. (I'm not sure if I've cleaned 
up lyx_main.C or path_defines.C.in as completely as I should.) LyX was 
reconfigured to add a version suffix, and the script is named "lyx". 
Patches attached.

I've also attached a copy of my shell scripts that set the path, 
userdir, and sysdir for lyx-13x for your review. (The one for lyx-140 
is pretty much the same.)

I've also put a copy of the resulting lyx-1.3.6 here: 
. Andreas, 
can you test that it works for you?

Bennett


lyx
Description: Binary data


path-140.patch
Description: Binary data


path-13x.patch
Description: Binary data

Re: LyX + MinGW + MinSYS + GPLed QT/Win32 --- success!

[Lars Gullik Bjønnes]

Angus Leeming <[EMAIL PROTECTED]> writes:

| Let's address each issue as I post a chunk. However, I have a general
| question:
>
| os_win32.h is a kludge. It is performing the same rôle as config.h. I think
| that these things should go in config.h, which suggests a config/win32.m4
| file.
| * Do you agree?

maybe.

| * Is it possible to add tests to configure that effect only Win32 users.
| Ie, config.h won't be changed as I add tests except on Win32 machines?

No. Make the tests general.

-- 
Lgb

Re: sideways tables/figures

[Jean-Marc Lasgouttes]

> "Juergen" == Juergen Spitzmueller <[EMAIL PROTECTED]> writes:

Juergen> Edwin Leuven wrote:
>> updated patch attached. have been using it the last couple of days
>> and have experienced no regressions.

Juergen> You have noticed that this feature also has been added to
Juergen> 1.4?

Juergen, concerning the sideways feature, do you know about the
rotfloat.sty package, which allows to rotate all kind of floats?

JMarc

Re: [patch] encoding error

[Juergen Spitzmueller]

Jean-Marc Lasgouttes wrote:

> Juergen> - "textrm Å", "textrm Ø", "mathcircumflex", "_",
> Juergen> + "textrm Å", "textrm Ø", "mathcircumflex", "_",
> 
> Is it really a good idea to have stuff in latin1 included
> unconditionally? Why not use "AA" and "emptyset"?

Yes, I agree, but I think that \textrm{Ø} is rather synonymical to
\textrm{\O} than to \textrm{\emptyset} (which also triggers a LaTeX error).
But OTOH I am no mathematician.

Also, mathed does not render \AA and \O to symbols ATM, but I guess André
could fix that easily.

Jürgen.

Re: LyX + MinGW + MinSYS + GPLed QT/Win32 --- success!

[Angus Leeming]

Jean-Marc Lasgouttes wrote:
> Do the math fonts work correctly? Ruurd once said they were too large.

I didn't have a document to try it out with last night. I got it to start
and that was more than enough. (It takes twenty minutes just to link the
bloody thing...)

> How come you do not need the dirent stuff?

Shrug. Dunno.
> 
> Angus> The resulting executable is rather neutered because the
> Angus> fork/execvp stuff is extremely ropey. It compiles, but the
> Angus> whole waitpid thing is guaranteed not to work ;-)
> 
> So this will be done later, right?

That's the idea.

> I guess the patch will need some cleanup before going in. Do you
> prefer to have remarks now or when the chunks come in?

Let's address each issue as I post a chunk. However, I have a general
question:

os_win32.h is a kludge. It is performing the same rôle as config.h. I think
that these things should go in config.h, which suggests a config/win32.m4
file.
* Do you agree?
* Is it possible to add tests to configure that effect only Win32 users.
Ie, config.h won't be changed as I add tests except on Win32 machines?

-- 
Angus

Equation crashes cvs-1.4.0 (12/12/04)

[Fred Nastos]

Hi,

I downloaded and built the cvs version last night.  Athough I can
start it up and create new documents, it seems to crash (without
starting up) on almost all my documents from 1.3.5.  I thought you
might want to see this, so I managed to pare one document down
to a single equation as an example, which I attach.  It seems like
it's the multiline equations that it doesn't like.

Thanks
#LyX 1.3 created this file. For more info see http://www.lyx.org/
\lyxformat 221
\textclass article
\begin_preamble
\usepackage{ae,aecompl}
\end_preamble
\language english
\inputencoding auto
\fontscheme default
\graphics default
\paperfontsize 11
\spacing single 
\papersize letterpaper
\paperpackage a4
\use_geometry 1
\use_amsmath 1
\use_natbib 0
\use_numerical_citations 0
\paperorientation portrait
\leftmargin 1in
\rightmargin 1in
\secnumdepth 3
\tocdepth 3
\paragraph_separation indent
\defskip medskip
\quotes_language english
\quotes_times 2
\papercolumns 1
\papersides 1
\paperpagestyle default

\layout Standard

This seems to break lyx cvs 1.4.0
\begin_inset Formula \begin{align*}
\int_{V}d^{3}r\,\psi_{n\vec{k}}^{*}(\vec{r})\psi_{m\vec{q}}(\vec{r}) & =\frac{\Omega}{8\pi^{3}}\int_{V}d^{3}r\, u_{n\vec{k}}^{*}(\vec{r})u_{m\vec{q}}(\vec{r})\mathrm{e}^{\mathrm{i}\left(\vec{q}-\vec{k}\right)\cdot\vec{r}},\\
 & =\frac{\Omega}{8\pi^{3}}\sum_{\mbox{unit cells}}\int_{\Omega}d^{3}r\, u_{n\vec{k}}^{*}(\vec{r}+\vec{R}_{n})u_{m\vec{q}}(\vec{r}+\vec{R}_{n})\mathrm{e}^{\mathrm{i}\left(\vec{q}-\vec{k}\right)\cdot\left(\vec{r}+\vec{R}_{n}\right)},\\
 & =\frac{\Omega}{8\pi^{3}}\sum_{\mbox{unit cells}}\int_{\Omega}d^{3}r\, u_{n\vec{k}}^{*}(\vec{r})u_{m\vec{q}}(\vec{r})\mathrm{e}^{\mathrm{i}\left(\vec{q}-\vec{k}\right)\cdot\left(\vec{r}+\vec{R}_{n}\right)},\\
 & =\frac{\Omega}{8\pi^{3}}\int_{\Omega}d^{3}r\, u_{n\vec{k}}^{*}(\vec{r})u_{m\vec{q}}(\vec{r})\mathrm{e}^{\mathrm{i}\left(\vec{q}-\vec{k}\right)\cdot\vec{r}}\sum_{\mbox{unit cells}}\mathrm{e}^{\mathrm{i}\left(\vec{q}-\vec{k}\right)\cdot\vec{R}_{n}},\\
 & =\frac{\Omega}{8\pi^{3}}\int_{\Omega}d^{3}r\, u_{n\vec{k}}^{*}(\vec{r})u_{m\vec{q}}(\vec{r})\mathrm{e}^{\mathrm{i}\left(\vec{q}-\vec{k}\right)\cdot\vec{r}}N\sum_{\vec{K}}\delta_{\vec{K},\vec{q}-\vec{k}},\\
 & =\frac{V}{8\pi^{3}}\int_{\Omega}d^{3}r\, u_{n\vec{k}}^{*}(\vec{r})u_{m\vec{q}}(\vec{r})\mathrm{e}^{\mathrm{i}\left(\vec{q}-\vec{k}\right)\cdot\vec{r}}\sum_{\vec{K}}\delta_{\vec{K},\vec{q}-\vec{k}},\\
 & =\frac{V}{8\pi^{3}}\int_{\Omega}d^{3}r\, u_{n\vec{k}}^{*}(\vec{r})u_{m\vec{q}}(\vec{r})\mathrm{e}^{\mathrm{i}\left(\vec{q}-\vec{k}\right)\cdot\vec{r}}\delta_{\vec{q},\vec{k}},\quad\mbox{where we have limited $\vec{q},$$\vec{k}\in FBZ.$}\\
 & =\frac{V}{8\pi^{3}}\delta_{\vec{q},\vec{k}}\delta_{n,m}.\end{align*}

\end_inset 


\the_end

Re: [patch] partial fix for bug 220

[Jean-Marc Lasgouttes]

> "Georg" == Georg Baum <[EMAIL PROTECTED]> writes:

Georg> The attached patch adds partial longtable support and some
Georg> small fixes to tex2lyx. This is needed for bug 220 (gnumeric
Georg> table import). Double hlines (firstHeadTopDL etc. in LyX) are
Georg> missing. The plan is to add this in a second step. That will
Georg> also fix double hlines in normal tables. I copied some structs
Georg> from tabular.h instead of including math_gridinset.h, because I
Georg> needed to modify them. I guess that these (and some common
Georg> code) could go to a common base class for tex2lyx table stuff
Georg> and LyXTabular when the table import is "finished", but for now
Georg> I don't want to touch tabular.C too much. I had to rearrange
Georg> the code quite a bit, but I think that it is understandable
Georg> now. OK to commit?

While I cannot comment much on whether it really works, it looks sound
to me. Go for it.

Remember too that it may be a good idea to create a test-tabular.tex
file to get some torture test for the code...

JMarc

Re: LyX + MinGW + MinSYS + GPLed QT/Win32 --- success!

[Jean-Marc Lasgouttes]

> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:

Angus> I spent a good chunk of this weekend trying to get LyX to
Angus> compile on Windows using the Qt port from the cygwin-kde people
Angus> and I've finally succeeded. 

Excellent!

Do the math fonts work correctly? Ruurd once said they were too large.

Angus> Attached are some notes detailing what I needed to do, together
Angus> with a patch and os_win32.h which should go in src/support. The
Angus> patch is essentially Ruurd's, but I've put it together piece by
Angus> piece as compilation failed.

How come you do not need the dirent stuff?

Angus> The resulting executable is rather neutered because the
Angus> fork/execvp stuff is extremely ropey. It compiles, but the
Angus> whole waitpid thing is guaranteed not to work ;-)

So this will be done later, right?

Angus> What I'd like to do is to break this patch into smaller parts
Angus> and apply each piece with a proper explanation about what it
Angus> does and why its needed. I've only compiled lyx-13x, but much
Angus> of the patch will apply without changes to the 14x tree too.

Angus> So, JMarc and Lars. Are you happy if I put together small
Angus> patches for the 13x and 14x trees and proceed SLOWLY ?

I am very happy about it, as long as they go in HEAD too.

I guess the patch will need some cleanup before going in. Do you
prefer to have remarks now or when the chunks come in?

JMarc

Re: updating from cvs: missing files or what?

[Jean-Marc Lasgouttes]

> "Hartmut" == Hartmut Haase <[EMAIL PROTECTED]> writes:

Hartmut> During the last couple of weeks I updated several times, but
Hartmut> make always failed. Today it stopped with the following
Hartmut> messages:

Hartmut> In SpecInfo
Hartmut> [fd_spec.c 247] Skipped object class 18(Input) In TypeValue
Hartmut> [fd_objects.c 629] type is unknown In SpecInfo [fd_spec.c
Hartmut> 247] Skipped object class 18(Input) In fl_bgn_group [util.c
Hartmut> 65] You forgot to call fl_end_group. In fl_bgn_form [forms.c
Hartmut> 79] You forgot to call fl_end_form "fdesign -convert
Hartmut> form_print.fd}" failed. Please investigate. 

The error does not appear here with the following version of fdesign:

fantomas: fdesign -version
fdesign (FORM Designer) Exp 1.9 of 2003/11/27 14:51:04
Copyright (c) 1996-2002 by T.C. Zhao and Mark Overmars

What is your version?

JMarc

[patch] credits

[Juergen Spitzmueller]

I know it's feature freeze, but this one should go in ;-)

Jürgen
Index: CREDITS
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/lib/CREDITS,v
retrieving revision 1.49
diff -u -r1.49 CREDITS
--- CREDITS	26 Oct 2004 21:16:41 -	1.49
+++ CREDITS	13 Dec 2004 15:06:58 -
@@ -260,6 +260,9 @@
 @bJürgen Spitzmüller
 @iE-mail: [EMAIL PROTECTED]
Qt frontend, bugfixes
[EMAIL PROTECTED] Spray
[EMAIL PROTECTED]: [EMAIL PROTECTED]
+   Gtk frontend
 @bBen Stanley
 @iE-mail: [EMAIL PROTECTED]
fix bugs with error insets placement

Re: [patch] encoding error

[Jean-Marc Lasgouttes]

> "Juergen" == Juergen Spitzmueller <[EMAIL PROTECTED]> writes:

Juergen> I don't know who introduced this, but it is breaking two
Juergen> functions in the math panel. I'll apply if I get no
Juergen> objections.

Juergen> - "textrm Å", "textrm Ø", "mathcircumflex", "_", 
Juergen> + "textrm Å", "textrm Ø", "mathcircumflex", "_", 

Is it really a good idea to have stuff in latin1 included
unconditionally? Why not use "AA" and "emptyset"?

JMarc

Re: [patch] credits

[Lars Gullik Bjønnes]

Juergen Spitzmueller <[EMAIL PROTECTED]> writes:

| I know it's feature freeze, but this one should go in ;-)

Yes.

-- 
Lgb

Re: [patch] credits

[Juergen Spitzmueller]

Lars Gullik Bjønnes wrote:
> | I know it's feature freeze, but this one should go in ;-)
>
> Yes.

Done.

Jürgen

Re: [patch] credits

[Alfredo Braunstein]

Juergen Spitzmueller wrote:

> I know it's feature freeze, but this one should go in ;-)

I think so also ;-)

Shove it in

Alfredo

[patch] encoding error

[Juergen Spitzmueller]

I don't know who introduced this, but it is breaking two functions in the math 
panel. I'll apply if I get no objections.

Jürgen
Index: ControlMath.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/controllers/ControlMath.C,v
retrieving revision 1.26
diff -u -r1.26 ControlMath.C
--- ControlMath.C	19 May 2004 15:11:30 -	1.26
+++ ControlMath.C	13 Dec 2004 15:01:57 -
@@ -177,7 +177,7 @@
 	"angle", "top", "bot", "Vert", "neg",
 	"flat", "natural", "sharp", "surd", "triangle",
 	"diamondsuit", "heartsuit", "clubsuit", "spadesuit",
-	"textrm Ã?", "textrm Ã?", "mathcircumflex", "_",
+	"textrm Å", "textrm Ø", "mathcircumflex", "_",
 	"mathrm T",
 	"mathbb N", "mathbb Z", "mathbb Q",
 	"mathbb R", "mathbb C", "mathbb H",

Re: sideways tables/figures

[Lars Gullik Bjønnes]

Edwin Leuven <[EMAIL PROTECTED]> writes:

>> Yes. And if you are willing to test the 1.4 version (and compare
>> with your stuff), we can improve the support in the 1.5 cycle :-)
>
| can 1.4cvs be safely used for real work now?

I'd still wait a bit, but it is a lot better than just a few weeks
ago.

Otoh we need testing...

-- 
Lgb

Re: sideways tables/figures

[Edwin Leuven]

Yes. And if you are willing to test the 1.4 version (and compare with your 
stuff), we can improve the support in the 1.5 cycle :-)
can 1.4cvs be safely used for real work now?
ed.

updating from cvs: missing files or what?

[Hartmut Haase]

During the last couple of weeks I updated several times, but make always 
failed. Today it stopped with the following messages:

Making all in forms
make[6]: Entering directory 
`/home/haase/Software/lyx-devel/src/frontends/xforms/forms'
{ [ . != . ] && [ ! -r form_print.fd ] && ln -s form_print.fd . ; } || true
/bin/sh ./fdfix.sh form_print.fd
In SpecInfo [fd_spec.c 247] Skipped object class 18(Input)
In TypeValue [fd_objects.c 629] type  is unknown
In SpecInfo [fd_spec.c 247] Skipped object class 18(Input)
In fl_bgn_group [util.c 65] You forgot to call fl_end_group.
In fl_bgn_form [forms.c 79] You forgot to call fl_end_form
"fdesign -convert form_print.fd}" failed. Please investigate.
make[6]: *** [form_print.C] Fehler 1
make[6]: Leaving directory 
`/home/haase/Software/lyx-devel/src/frontends/xforms/forms'
...
anyone any ideas what that means?
-- 
Hartmut 

Hungerhilfe: http://www.thehungersite.com

Das heutige Motto:
Es gibt mehr Dinge zwischen Himmel und Erde als Deine Schulweisheit Dich 
träumen läßt.

Re: sideways tables/figures

[Juergen Spitzmueller]

Edwin Leuven wrote:
> nope. so the transfer will be smooth ;-)

Yes. And if you are willing to test the 1.4 version (and compare with your 
stuff), we can improve the support in the 1.5 cycle :-)

Jürgen 

Re: sideways tables/figures

[Edwin Leuven]

You have noticed that this feature also has been added to 1.4?
nope. so the transfer will be smooth ;-)
regards, ed.

Re: sideways tables/figures

[Juergen Spitzmueller]

Edwin Leuven wrote:
> updated patch attached. have been using it the last couple of days and
> have experienced no regressions.

You have noticed that this feature also has been added to 1.4?

Jürgen

Re: sideways tables/figures

[Edwin Leuven]

updated patch attached. have been using it the last couple of days and 
have experienced no regressions.

edwin


sideways.diff.gz
Description: application/gzip