Re: [Rpcemu] RPCEmu and Networking

In article <c065689d54.Bernard@bapfish.plus.com>,
Bernard <bernard@bapfish.org.uk> wrote:
> In message <549d342cb7dave@triffid.co.uk>
> Dave Symes <dave@triffid.co.uk> wrote:

> > On 27 Feb, dave@triffid.co.uk wrote:
> >> RPCEmu 0.8.12 on RO 6.20 and for most of the time it network connects
> >> via the bridge okay.

> >> This problem has been occurring for some time and I've just let it
> >> pass, however, it has happened twice today and is annoying me...

> >> Q1) Anyone have ideas about what's going on?

> >> Q2) Thoughts on a fix other than a reboot?

> >> Thanks
> >> Dave

> > And it ********** happened again a short while ago.
> > D.

> Get yourself a Raspberry Pi.

> Bernard

Thank you Bernard.
Unfortunately ATM. I don't have the spare time to play with toy computers
as I'm a bit busy. ;-)

Dave

--

Dave Triffid

_______________________________________________
Rpcemu mailing list
Rpcemu@riscos.info
http://www.riscos.info/cgi-bin/mailman/listinfo/rpcemu

Re: [Rpcemu] RPCEmu and Networking

In message <549d342cb7dave@triffid.co.uk>
Dave Symes <dave@triffid.co.uk> wrote:

> On 27 Feb, dave@triffid.co.uk wrote:
>> RPCEmu 0.8.12 on RO 6.20 and for most of the time it network connects
>> via the bridge okay.

>> This problem has been occurring for some time and I've just let it pass,
>> however, it has happened twice today and is annoying me...

>> Q1) Anyone have ideas about what's going on?

>> Q2) Thoughts on a fix other than a reboot?

>> Thanks
>> Dave

> And it ********** happened again a short while ago.
> D.

Get yourself a Raspberry Pi.

Bernard
--
Contact Me - http://www.bapfish.org.uk/contactme/
Web - http://www.bapfish.org.uk/

_______________________________________________
Rpcemu mailing list
Rpcemu@riscos.info
http://www.riscos.info/cgi-bin/mailman/listinfo/rpcemu

[gccsdk] build for gcc4.1.2 fails too

Failure building gcc4.1.2

details below.

I haven't yet found the solution to this..  Any thoughts?

thanks

John

cd .. \
  && CONFIG_FILES=po/Makefile.in:po/Make-in \
     CONFIG_HEADERS= /bin/bash ./config.status
config.status: creating po/Makefile.in
config.status: executing depfiles commands
config.status: executing libtool commands
config.status: executing default-1 commands
config.status: executing bfd_stdint.h commands
config.status: executing default commands
make[5]: Nothing to be done for 'info'.
make[5]: Leaving directory '/home/jb/GCCSDK412/gcc4/builddir/cross-binutils/bfd/po'
make[5]: Entering directory '/home/jb/GCCSDK412/gcc4/builddir/cross-binutils/bfd'
make[5]: Nothing to be done for 'info-am'.
make[5]: Leaving directory '/home/jb/GCCSDK412/gcc4/builddir/cross-binutils/bfd'
Makefile:1552: recipe for target 'info-recursive' failed
make[4]: *** [info-recursive] Error 1
make[4]: Leaving directory '/home/jb/GCCSDK412/gcc4/builddir/cross-binutils/bfd'
Makefile:3333: recipe for target 'all-bfd' failed
make[3]: *** [all-bfd] Error 2
make[3]: Leaving directory '/home/jb/GCCSDK412/gcc4/builddir/cross-binutils'
Makefile:851: recipe for target 'all' failed
make[2]: *** [all] Error 2
make[2]: Leaving directory '/home/jb/GCCSDK412/gcc4/builddir/cross-binutils'
Makefile:166: recipe for target 'buildstepsdir/cross-binutils' failed
make[1]: *** [buildstepsdir/cross-binutils] Error 2
make[1]: Leaving directory '/home/jb/GCCSDK412/gcc4'
Makefile:131: recipe for target 'getenv' failed
make: *** [getenv] Error 2

--

Stop press! Just announced -

Free Standard Broadband for 6 months, unlimited data, and no minimum contract

For full details of see http://www.JustAskJB.co.uk

Intrigued? Call me

John Ballance C.Eng MIET - jwb@macpcrepair.co.uk - 07976 295923

[gccsdk] patch needed in gcc4.7.1 build system

Hi

Trying to build gcc 4.7.1 build system falls over (see below). There is a patch that can fix this (see https://sourceware.org/ml/crossgcc/2014-05/msg00032.html  included below).

Can/should that go back in, or do I just apply it locally to resolve.  (basically there are 4 source files with changes of ptrdiff_t to std::ptrdiff_t, and inclusion of <cstddef>)

Thanks
John

fallover:-

libtool: compile:  g++ -DHAVE_CONFIG_H -I. -I/home/jb/GCCSDK471/gcc4/srcdir/ppl/src -I.. -I.. -I/home/jb/GCCSDK471/gcc4/srcdir/ppl/src -I/home/jb/GCCSDK471/gcc4/builddir/installed-libs-for-cross-gcc/include -g -O2 -frounding-math -W -Wall -MT Grid_Generator.lo -MD -MP -MF .deps/Grid_Generator.Tpo -c /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.cc -o Grid_Generator.o
In file included from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator_System.defs.hh:28:0,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.defs.hh:29,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.cc:26:
/home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Generator_System.defs.hh:253:7: error: 'ptrdiff_t' was not declared in this scope
       ptrdiff_t,
       ^
/home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Generator_System.defs.hh:253:7: note: suggested alternatives:
In file included from /usr/include/c++/4.9/iostream:38:0,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/checked.defs.hh:27,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Checked_Number.defs.hh:28,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Coefficient.types.hh:16,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Coefficient.defs.hh:27,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.defs.hh:28,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.cc:26:
/usr/include/i386-linux-gnu/c++/4.9/bits/c++config.h:189:28: note:   'std::ptrdiff_t'
   typedef __PTRDIFF_TYPE__ ptrdiff_t;
                            ^
/usr/include/i386-linux-gnu/c++/4.9/bits/c++config.h:189:28: note:   'std::ptrdiff_t'
In file included from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator_System.defs.hh:28:0,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.defs.hh:29,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.cc:26:
/home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Generator_System.defs.hh:255:23: error: template argument 3 is invalid
       const Generator&> {
                       ^
In file included from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.defs.hh:29:0,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.cc:26:
/home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator_System.defs.hh:270:7: error: 'ptrdiff_t' was not declared in this scope
       ptrdiff_t,
       ^
/home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator_System.defs.hh:270:7: note: suggested alternatives:
In file included from /usr/include/c++/4.9/iostream:38:0,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/checked.defs.hh:27,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Checked_Number.defs.hh:28,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Coefficient.types.hh:16,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Coefficient.defs.hh:27,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.defs.hh:28,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.cc:26:
/usr/include/i386-linux-gnu/c++/4.9/bits/c++config.h:189:28: note:   'std::ptrdiff_t'
   typedef __PTRDIFF_TYPE__ ptrdiff_t;
                            ^
/usr/include/i386-linux-gnu/c++/4.9/bits/c++config.h:189:28: note:   'std::ptrdiff_t'
In file included from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.defs.hh:29:0,
                 from /home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator.cc:26:
/home/jb/GCCSDK471/gcc4/srcdir/ppl/src/Grid_Generator_System.defs.hh:272:28: error: template argument 3 is invalid
       const Grid_Generator&>,
                            ^
Makefile:1159: recipe for target 'Grid_Generator.lo' failed
make[5]: *** [Grid_Generator.lo] Error 1
make[5]: Leaving directory '/home/jb/GCCSDK471/gcc4/builddir/cross-ppl/src'
Makefile:951: recipe for target 'all' failed
make[4]: *** [all] Error 2
make[4]: Leaving directory '/home/jb/GCCSDK471/gcc4/builddir/cross-ppl/src'
Makefile:478: recipe for target 'all-recursive' failed
make[3]: *** [all-recursive] Error 1
make[3]: Leaving directory '/home/jb/GCCSDK471/gcc4/builddir/cross-ppl'
Makefile:405: recipe for target 'all' failed
make[2]: *** [all] Error 2
make[2]: Leaving directory '/home/jb/GCCSDK471/gcc4/builddir/cross-ppl'
Makefile:424: recipe for target 'cross-ppl-built' failed
make[1]: *** [cross-ppl-built] Error 2
make[1]: Leaving directory '/home/jb/GCCSDK471/gcc4'
Makefile:252: recipe for target 'getenv' failed
make: *** [getenv] Error 2


the patch:-

# HG changeset patch  # User Bernhard Walle <bernhard@bwalle.de>  # Date 1399806036 -7200  #      Sun May 11 13:00:36 2014 +0200  # Node ID 5632a171e7c1dc4c0fb5c7a8352afb4188cd29d5  # Parent  e35fa03cd2043496839aa5fa3a1a3d2d375b5535  complibs/ppl: Fix build of ppl 0.11.2 with gcc 4.9      From: Roberto Bagnara <bagnara@cs.unipr.it>    Date: Tue, 29 Apr 2014 19:51:43 +0000 (+0200)    Subject: Added missing inclusions.  Use std::ptrdiff_t.    X-Git-Url: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl%2Fppl.git;a=commitdiff_plain;h=61d4e14dfd9f1121e9b4521dead5728b2424dd7c      Added missing inclusions.  Use std::ptrdiff_t.    (Thanks to Paulo Cesar Pereira de Andrade.)    Signed-off-by: Bernhard Walle <bernhard@bwalle.de>    diff -r e35fa03cd204 -r 5632a171e7c1 patches/ppl/0.11.2/500-ptrdiff_t.patch  --- /dev/null	Thu Jan 01 00:00:00 1970 +0000  +++ b/patches/ppl/0.11.2/500-ptrdiff_t.patch	Sun May 11 13:00:36 2014 +0200  @@ -0,0 +1,95 @@  +From: Bernhard Walle <bernhard@bwalle.de>  +Subject: Fix compilation with gcc 4.9 (ptrdiff_t errors)  +  +This is a backport of following patch in ppl git  +  +  commit 61d4e14dfd9f1121e9b4521dead5728b2424dd7c  +  Author: Roberto Bagnara <bagnara@cs.unipr.it>  +  Date:   Tue Apr 29 21:51:43 2014 +0200  +  +      Added missing inclusions.  Use std::ptrdiff_t.  +      (Thanks to Paulo Cesar Pereira de Andrade.)  +  +---  + src/Congruence_System.defs.hh     |    3 ++-  + src/Constraint_System.defs.hh     |    3 ++-  + src/Generator_System.defs.hh      |    3 ++-  + src/Grid_Generator_System.defs.hh |    3 ++-  + 4 files changed, 8 insertions(+), 4 deletions(-)  +  +--- a/src/Grid_Generator_System.defs.hh  ++++ b/src/Grid_Generator_System.defs.hh  +@@ -30,6 +30,7 @@ site: http://www.cs.unipr.it/ppl/ . */  + #include "Variables_Set.types.hh"  + #include "Grid.types.hh"  + #include <iosfwd>  ++#include <cstddef>  +   + namespace Parma_Polyhedra_Library {  +   +@@ -267,7 +268,7 @@ public:  +   class const_iterator  +     : public std::iterator<std::forward_iterator_tag,  + 			   Grid_Generator,  +-			   ptrdiff_t,  ++			   std::ptrdiff_t,  + 			   const Grid_Generator*,  + 			   const Grid_Generator&>,  +       private Generator_System::const_iterator {  +--- a/src/Congruence_System.defs.hh  ++++ b/src/Congruence_System.defs.hh  +@@ -33,6 +33,7 @@ site: http://www.cs.unipr.it/ppl/ . */  + #include "Grid.types.hh"  + #include "Grid_Certificate.types.hh"  + #include <iosfwd>  ++#include <cstddef>  +   + namespace Parma_Polyhedra_Library {  +   +@@ -235,7 +236,7 @@ public:  +   class const_iterator  +     : public std::iterator<std::forward_iterator_tag,  + 			   Congruence,  +-			   ptrdiff_t,  ++			   std::ptrdiff_t,  + 			   const Congruence*,  + 			   const Congruence&> {  +   public:  +--- a/src/Constraint_System.defs.hh  ++++ b/src/Constraint_System.defs.hh  +@@ -33,6 +33,7 @@ site: http://www.cs.unipr.it/ppl/ . */  + #include "Congruence_System.types.hh"  + #include <iterator>  + #include <iosfwd>  ++#include <cstddef>  +   + namespace Parma_Polyhedra_Library {  +   +@@ -204,7 +205,7 @@ public:  +   class const_iterator  +     : public std::iterator<std::forward_iterator_tag,  + 			   Constraint,  +-			   ptrdiff_t,  ++			   std::ptrdiff_t,  + 			   const Constraint*,  + 			   const Constraint&> {  +   public:  +--- a/src/Generator_System.defs.hh  ++++ b/src/Generator_System.defs.hh  +@@ -33,6 +33,7 @@ site: http://www.cs.unipr.it/ppl/ . */  + #include "Polyhedron.types.hh"  + #include "Poly_Con_Relation.defs.hh"  + #include <iosfwd>  ++#include <cstddef>  +   + namespace Parma_Polyhedra_Library {  +   +@@ -250,7 +251,7 @@ public:  +   class const_iterator  +     : public std::iterator<std::forward_iterator_tag,  + 			   Generator,  +-			   ptrdiff_t,  ++			   std::ptrdiff_t,  + 			   const Generator*,  + 			   const Generator&> {  +   public:    --  

--

Stop press! Just announced -

Free Standard Broadband for 6 months, unlimited data, and no minimum contract

For full details of see http://www.JustAskJB.co.uk

Intrigued? Call me

John Ballance C.Eng MIET - jwb@macpcrepair.co.uk - 07976 295923

Re: [Rpcemu] RPCEmu and Networking

On 27 Feb, dave@triffid.co.uk wrote:
> RPCEmu 0.8.12 on RO 6.20 and for most of the time it network connects
> via the bridge okay.

> This problem has been occurring for some time and I've just let it pass,
> however, it has happened twice today and is annoying me...

> Regardless of which PC I'm running a copy of RPCEmu 0.8.12 and RO 6.20
> on, there are times when after running, it just will not connect to the
> net.

> The Mail or News app is trying to Resolve, but there just ain't a
> connection.

> If I now turn to my Real SARPC also running RO 6.20 I can immediately
> connect to the network.

> Turn back to the RPCEmu and it still can't Resolve, nor Browse. Dropping
> out of Emu to the Windows side and the network is working okay... Back
> into Emu and no connection.

> The only fix is to completely closed down RPCEmu, then run it again.

> As noted at the beginning, mostly the bridge works okay, but a few times
> a week, it will not connect, even though the RO computer next to it is
> showing the network connection through the router is okay, as it is for
> the Win side.

> Q1) Anyone have ideas about what's going on?

> Q2) Thoughts on a fix other than a reboot?

> Thanks
> Dave

And it ********** happened again a short while ago.
D.

--

Dave Triffid

_______________________________________________
Rpcemu mailing list
Rpcemu@riscos.info
http://www.riscos.info/cgi-bin/mailman/listinfo/rpcemu

[gccsdk] Building with autobuilder

Hi

I'm trying to build some of the autobuilder targets, including hopefully firefox, but I keep falling over fail-to-build issues.

Next issue.. trying to build with the autobuilder, gettext fails.

in build:   ../autobuilder/build-libs just reloaded the sources again and overrode my mods.

it uses gettext_0.19.3 updated with gettext_0.19.3-2.debian. ...

..and is it that my arm-unknown-riscos/include/sys/time.h has built wrongly or have things just moved 'out of sight' again?

error stuff around here:

/home/jb/GCCSDK/env/arm-unknown-riscos-gcc -DHAVE_CONFIG_H -DEXEEXT=\",e1f\" -I. -I..  -I../intl -I../intl -DDEPENDS_ON_LIBICONV=1 -DDEPENDS_ON_LIBINTL=1    -O3 -c -o gettimeofday.o gettimeofday.c
In file included from ./sys/time.h:39:0,
                 from gettimeofday.c:23:
/home/jb/GCCSDK/cross/lib/gcc/arm-unknown-riscos/4.7.4/../../../../arm-unknown-riscos/include/sys/time.h:74:12: error: conflicting types for 'rpl_gettimeofday'
In file included from /home/jb/GCCSDK/cross/lib/gcc/arm-unknown-riscos/4.7.4/../../../../arm-unknown-riscos/include/sys/select.h:22:0,
                 from /home/jb/GCCSDK/cross/lib/gcc/arm-unknown-riscos/4.7.4/../../../../arm-unknown-riscos/include/sys/time.h:30,
                 from ./sys/time.h:39,
                 from gettimeofday.c:23:
./sys/time.h:407:1: note: previous declaration of 'rpl_gettimeofday' was here
Makefile:1460: recipe for target 'gettimeofday.o' failed

How does one oblige this to build with mods?

Thanks

John

--

Stop press! Just announced -

Free Standard Broadband for 6 months, unlimited data, and no minimum contract

For full details of see http://www.JustAskJB.co.uk

Intrigued? Call me

John Ballance C.Eng MIET - jwb@macpcrepair.co.uk - 07976 295923

3.3 release, last call for fixes

Any developers with important fixes please commit them. If you have a
reason to call a hold on the release please also yell RSN.

Bug tracker has 2270 and 2228 tagged for this release, I intend to
push these tags to 3.4 instead

--
Regards Vincent
http://www.kyllikki.org/

Re: [gccsdk] Building the GCCSDK - current status?

In message <54EF8EF2.60607@macpcrepair.co.uk>
John Ballance <jwb@macpcrepair.co.uk> wrote:

> On 26/02/2015 19:17, Lee Noar wrote:
> > On 26/02/15 16:13, John Ballance wrote:
> >> Hi Folks
> >>
> >> I thought it was time to look at this. I've followed the various
> >> instructions at http://www.riscos.info/index.php/Using_GCCSDK and more
> >> or less got a working crossbuild environment. I did gave to patch ppl
> >> for a ptrdiff_t error.
> >
> > I haven't seen this problem myself; perhaps an issue with the native GCC
>
> native gcc is 4.9.1 on ubuntu .. is that as expected?

More than recent enough.

> > you're using to build the cross compiler? I'm using v4.8.2. I've just
> > recently rebuilt the 4.7.4 compiler from scratch and the only issue was
> > with downloading the source for GMP which gave this:
> >
> > --2015-02-26 18:46:54--
> > ftp://ftp.gmplib.org/pub/gmp-5.0.1/gmp-5.0.1.tar.bz2
> > => 'gmp-5.0.1.tar.bz2'
> > Resolving ftp.gmplib.org (ftp.gmplib.org)... 130.242.124.100
> > Connecting to ftp.gmplib.org (ftp.gmplib.org)|130.242.124.100|:21...
> > connected.
> > Logging in as anonymous ... Logged in!
> > ==> SYST ... done. ==> PWD ... done.
> > ==> TYPE I ... done. ==> CWD (1) /pub/gmp-5.0.1 ... done.
> > ==> SIZE gmp-5.0.1.tar.bz2 ... 2006109
> > ==> PASV ...
>
> What I did was to ftp the relevant file from the site it was trying and
> put it in the source.orig folder in gcc4.. the build then proceeded.

That's fine as workaround.

I've updated the GMP download URL with one which works. I'm not sure
why the previous one didn't work. So this problem should be solved for
any first time builders (or after a 'make distclean').

John.
--
John Tytgat, in his comfy chair at home
John.Tytgat@aaug.net

_______________________________________________
GCCSDK mailing list gcc@gccsdk.riscos.info
Bugzilla: http://www.riscos.info/bugzilla/index.cgi
List Info: http://www.riscos.info/mailman/listinfo/gcc
Main Page: http://www.riscos.info/index.php/GCCSDK

[Rpcemu] RPCEmu and Networking

RPCEmu 0.8.12 on RO 6.20 and for most of the time it network connects via
the bridge okay.

This problem has been occurring for some time and I've just let it pass,
however, it has happened twice today and is annoying me...

Regardless of which PC I'm running a copy of RPCEmu 0.8.12 and RO 6.20 on,
there are times when after running, it just will not connect to the net.

The Mail or News app is trying to Resolve, but there just ain't a
connection.

If I now turn to my Real SARPC also running RO 6.20 I can immediately
connect to the network.

Turn back to the RPCEmu and it still can't Resolve, nor Browse.
Dropping out of Emu to the Windows side and the network is working okay...
Back into Emu and no connection.

The only fix is to completely closed down RPCEmu, then run it again.

As noted at the beginning, mostly the bridge works okay, but a few times a
week, it will not connect, even though the RO computer next to it is
showing the network connection through the router is okay, as it is for
the Win side.

Q1) Anyone have ideas about what's going on?

Q2) Thoughts on a fix other than a reboot?

Thanks
Dave

--

Dave Triffid

_______________________________________________
Rpcemu mailing list
Rpcemu@riscos.info
http://www.riscos.info/cgi-bin/mailman/listinfo/rpcemu

Re: [gccsdk] alignment error in firefox2

On Thu, Feb 26, 2015 at 10:22:59PM +0000, John Ballance wrote:
> Hi
>
> Testing firefox2 (the zip download, not a local build) on the imx
> platform shows an alignment error
>
> The stderr o/p is included. Being unfamiliar with this backtrace I
> can see the issue, but would appreciate pointing to the code that is
> a bother.

That's pretty early on - it starts its version of main() and then starts
mucking about with its environment variables, at which point it triggers an
alignment exception. That looks like it ends up being a signal that gets
raised, at which point UnixLib's trap handler gets called and gives up.
(Somebody who knows more than me please correct me here)

I'd guess Firefox is built with unaligned loads, which would be the cause of
the problem.

Building Firefox is awkward because:
Versions of Firefox >=3 dropped the Xlib backend that uses ChoX11 on RISC OS
- so we're stuck with Firefox 2 unless someone ports a modern rendering
library
Firefox has a huge list of dependencies, many of which themselves have
moved on for one reason or another.
Not only may those dependencies not build, once built they could be
incompatible with FF 2 in some way (that the FF folks have zero interest in
fixing)

So a build means getting the exact versions of libraries contemporary with
FF 2 and building those in turn. That's quite awkward and difficult to
maintain. There could be some merit in using upstream sources from Ubuntu
9.04 or whatever, but it might need a dedicated autobuilder setup - and
might have to backdate the autobuilder checkout to match.

Theo

_______________________________________________
GCCSDK mailing list gcc@gccsdk.riscos.info
Bugzilla: http://www.riscos.info/bugzilla/index.cgi
List Info: http://www.riscos.info/mailman/listinfo/gcc
Main Page: http://www.riscos.info/index.php/GCCSDK

[gccsdk] alignment error in firefox2

Hi

Testing firefox2 (the zip download, not a local build) on the imx platform shows an alignment error

The stderr o/p is included. Being unfamiliar with this backtrace I can see the issue, but would appreciate pointing to the code that is a bother.

the assembler o/p below shows the error at ldr r12,[r0,#0], and r0 (a1) is cb1071bb, which is not word aligned.. hence the problem.

As a general rule the imx is run grumbling about word alignment exceptions, so I'd like to get to the bottom of this.

Any pointers to better understanding this error dump would be appreciated. (firefox reports as 2.0.0.22(pre))

Thanks
John





Fatal signal received: Segmentation fault
Stack backtrace:
Running thread 0x11e1540
  ( 138ff38) pc:   eacb2c lr:   ead140 sp:  138ff3c  __write_backtrace()
  ( 138ffa0) pc:   eacce4 lr:   ead990 sp:  138ffa4  __unixlib_raise_signal()
  ( 138ffb0) pc:   ead894 lr:   ecc288 sp:  138ec68  __h_cback()
  Register dump at 0138ffb4:
    a1: cb1071bb a2:  138ec68 a3:        1 a4:        0
    v1: cb1071d8 v2: cb1071bb v3: cb1071a0 v4: cb1071ba
    v5: cb107190 v6:       1a sl:  138e210 fp:  138eca4
    ip:  138eca8 sp:  138ec68 lr:   ecc288 pc:   ec50a0
    cpsr: 20000110
  00ec508c : .... : 00000000 : ANDEQ   R0,R0,R0
  00ec5090 : .... : 00000000 : ANDEQ   R0,R0,R0
  00ec5094 : .... : 00000000 : ANDEQ   R0,R0,R0
  00ec5098 : .... : 00000000 : ANDEQ   R0,R0,R0
  00ec509c : .... : 00000000 : ANDEQ   R0,R0,R0
  00ec50a0 : .Àå : e590c000 : LDR     R12,[R0,#0]
  00ec50a4 : .0 ã : e3a03001 : MOV     R3,#1
  00ec50a8 : .. á : e1a01000 : MOV     R1,R0
  00ec50ac : .4ƒá : e1833403 : ORR     R3,R3,R3,LSL #8
  ( 138eca4) pc:   ecc230 lr:   e7f7ac sp:  138eca8  __addenv_to_os()
  ( 138ecc8) pc:   e7f718 lr:   ec3824 sp:  138eccc  putenv()
  ( 138ece4) pc:   140c84 lr:   ec3824 sp:  138ece8  PR_SetEnv()
  ( 138eff0) pc:     a9d0 lr:   ecb7c0 sp:  138eff4  XRE_main()

--

Stop press! Just announced -

Free Standard Broadband for 6 months, unlimited data, and no minimum contract

For full details of see http://www.JustAskJB.co.uk

Intrigued? Call me

John Ballance C.Eng MIET - jwb@macpcrepair.co.uk - 07976 295923

Re: [gccsdk] Building the GCCSDK - current status?

On 26/02/2015 21:30, Lee Noar wrote:

On 26/02/15 20:10, John Ballance wrote:

Hi
On 26/02/2015 19:17, Lee Noar wrote:

On 26/02/15 16:13, John Ballance wrote:

Hi Folks

I thought it was time to look at this. I've followed the various
instructions at http://www.riscos.info/index.php/Using_GCCSDK and more
or less got a working crossbuild environment. I did gave to patch ppl
for a ptrdiff_t error.

I haven't seen this problem myself; perhaps an issue with the native GCC

native gcc is 4.9.1 on ubuntu .. is that as expected?

Yes, I can't see why v4.9.1 should be a problem.

If any packages that you need fail to build, then it's probably best to
report them here on the mail list and I'm sure someone will have a look
for you.

I'm usually happy to patch.. the Q really is what path to feed back the
accumulated knowledge

Oh, I see, well as Theo says, post to the list or create a ticket in
GCCSDK bugzilla which will send a message to the list.

thanks

A third Q: SharedUnixLibrary: who currently handles this?.. are the
sources accessible?

This is part of the Unixlib source and can be found at:

gcc4/srcdir/gcc/libunixlib/module/sul.s

Ok, Thanks.  To build it, is there a simple command?

If you've already built the cross compiler, then the binary can be found
at either:

gcc4/builddir/cross-gcc/arm-unknown-riscos/libunixlib/sul

or:

cross/bin/arm-unknown-riscos-sul

If you make changes to the source and want to rebuild it, then the
quickest way is to change to directory:

gcc4/builddir/cross-gcc/arm-unknown-riscos/libunixlib

and then "make install"

excellent
Cheers

John

Lee.

--

Stop press! Just announced -

Free Standard Broadband for 6 months, unlimited data, and no minimum contract

For full details of see http://www.JustAskJB.co.uk

Intrigued? Call me

John Ballance C.Eng MIET - jwb@macpcrepair.co.uk - 07976 295923

Re: [gccsdk] Building the GCCSDK - current status?

Hi

On 26/02/2015 19:17, Lee Noar wrote:

On 26/02/15 16:13, John Ballance wrote:

Hi Folks

I thought it was time to look at this. I've followed the various
instructions at http://www.riscos.info/index.php/Using_GCCSDK and more
or less got a working crossbuild environment. I did gave to patch ppl
for a ptrdiff_t error.

I haven't seen this problem myself; perhaps an issue with the native GCC

native gcc is 4.9.1 on ubuntu .. is that as expected?

you're using to build the cross compiler? I'm using v4.8.2. I've just
recently rebuilt the 4.7.4 compiler from scratch and the only issue was
with downloading the source for GMP which gave this:

--2015-02-26 18:46:54-- ftp://ftp.gmplib.org/pub/gmp-5.0.1/gmp-5.0.1.tar.bz2
           => 'gmp-5.0.1.tar.bz2'
Resolving ftp.gmplib.org (ftp.gmplib.org)... 130.242.124.100
Connecting to ftp.gmplib.org (ftp.gmplib.org)|130.242.124.100|:21... connected.
Logging in as anonymous ... Logged in!
==> SYST ... done.    ==> PWD ... done.
==> TYPE I ... done.  ==> CWD (1) /pub/gmp-5.0.1 ... done.
==> SIZE gmp-5.0.1.tar.bz2 ... 2006109
==> PASV ...

What I did was to ftp the relevant file from the site it was trying and put it in the source.orig folder in gcc4.. the build then proceeded.

It got stuck here, I had to CTRL-C to exit.

Anyone else see this, or is it just me?

I had to change the URL locally in the top level makefile to:

http://ftp.gnu.org/gnu/gmp/gmp-$(GMP_VERSION).tar.bz2

Attempts then to use the autobuilder bring up
frequent errors however.

That leads me to ask: is there a more accurate setup script, or has time
just taken its toll?

Some packages always target the latest source and it's inevitable that
the patches will become outdated causing patching at build time to fail.
OTOH, some packages target a specific version of their source, and the
patches stay stable.

Also, if patches are found needed, just to build what in theory should
build, what is the correct feedback path to use?

If any packages that you need fail to build, then it's probably best to
report them here on the mail list and I'm sure someone will have a look
for you.

I'm usually happy to patch.. the Q really is what path to feed back the accumulated knowledge

A third Q: SharedUnixLibrary: who currently handles this?.. are the
sources accessible?

This is part of the Unixlib source and can be found at:

gcc4/srcdir/gcc/libunixlib/module/sul.s

Ok, Thanks.  To build it, is there a simple command?

Lee.


_______________________________________________
GCCSDK mailing list gcc@gccsdk.riscos.info
Bugzilla: http://www.riscos.info/bugzilla/index.cgi
List Info: http://www.riscos.info/mailman/listinfo/gcc
Main Page: http://www.riscos.info/index.php/GCCSDK

--

Stop press! Just announced -

Free Standard Broadband for 6 months, unlimited data, and no minimum contract

For full details of see http://www.JustAskJB.co.uk

Intrigued? Call me

John Ballance C.Eng MIET - jwb@macpcrepair.co.uk - 07976 295923

Re: [gccsdk] Flex and Bison broken?

On Fri, 20 Feb 2015 12:37:33 -0000, alan buckley <alan_baa@hotmail.com>
wrote:

>> Ron wrote on Monday, February 16, 2015 10:19 PM:
>
>> In message <DUB120-W387DD353C097BCD506734EF0260@phx.gbl>
>> alan buckley <alan_baa@hotmail.com> wrote:
>
>> > WPB wrote on Date: Sat, 7 Feb 2015 20:27:31:
>> > > Guess what? I'm back!
>> > >
>> > > So, now with everyone's help (thank you), I've progressed to the
>> point
>> > > that 'make' is trying to invoke flex and then bison.
>> > >
>> > > Both fail.
[snip]
>
> I finally got a chance to try a rebuild of m4 and bison and unfortunately
> both need their autobuilder recipes updated for the latest
> version/compiler
> so won't build.
>
> Though having re-read the thread it seems your initial problem was with
> Make. Have you solved this with the variables you have found? Or does
> Make need to look for things on RISC OS in a more intelligent way?
>
> Have you some simple Makefile(s) and other source files for the other
> tools so I can try them out and see what is happening myself?
>
> I've never used m4, flex or bison in any details, but a few simple
> examples that show what isn't working would help.
>
> Regards,
> Alan

Sorry to go quiet on you - I got tired of fighting it for a while, and
gave up. Then when you sent this reply to the list, I tried again, but got
even more weird errors, and went off in a huff! ;)

I need to revisit it and see if any of Ron's ideas will fix things.

As for Make, I see Duncan's version as much superior to the one in the AB,
but I'm still finding it really hard to get all the tools to play nicely
together with sfix switching in place.

I'll try to get back to it soon and report back again.

Thanks for your input.

WPB

_______________________________________________
GCCSDK mailing list gcc@gccsdk.riscos.info
Bugzilla: http://www.riscos.info/bugzilla/index.cgi
List Info: http://www.riscos.info/mailman/listinfo/gcc
Main Page: http://www.riscos.info/index.php/GCCSDK

Re: [gccsdk] Building the GCCSDK - current status?

On Thu, Feb 26, 2015 at 04:13:22PM +0000, John Ballance wrote:
> Hi Folks
>
> I thought it was time to look at this. I've followed the various
> instructions at http://www.riscos.info/index.php/Using_GCCSDK and
> more or less got a working crossbuild environment. I did gave to
> patch ppl for a ptrdiff_t error. Attempts then to use the
> autobuilder bring up frequent errors however.
>
> That leads me to ask: is there a more accurate setup script, or has
> time just taken its toll?

You didn't tell us what distro you were using: because the autobuilder uses
a lot of system-wide tools, it's quite dependent on the distro you're
running on. This could probably be better handled, but the least-pain
option is to run in a VM of Debian or Ubuntu if that isn't your native
platform. (I'd suggest Ubuntu 14.04 as a sensible baseline these days)

Some autobuilder packages are broken, and Alan's been working through them
fixing them up. Upstream changes means it's always a moving target to some
degree.

> Also, if patches are found needed, just to build what in theory
> should build, what is the correct feedback path to use?

Post them to the list, or create a ticket in GCCSDK bugzilla (which will
turn into a post to the list).

> A third Q: SharedUnixLibrary: who currently handles this?.. are the
> sources accessible?

They're in SVN:
http://www.riscos.info/websvn/filedetails.php?repname=gccsdk&path=%2Ftrunk%2Fgcc4%2Frecipe%2Ffiles%2Fgcc%2Flibunixlib%2Fmodule%2Fsul.s

This list is the best place to raise any concerns...

Theo

_______________________________________________
GCCSDK mailing list gcc@gccsdk.riscos.info
Bugzilla: http://www.riscos.info/bugzilla/index.cgi
List Info: http://www.riscos.info/mailman/listinfo/gcc
Main Page: http://www.riscos.info/index.php/GCCSDK

Re: [gccsdk] Building the GCCSDK - current status?

On 26/02/15 16:13, John Ballance wrote:
> Hi Folks
>
> I thought it was time to look at this. I've followed the various
> instructions at http://www.riscos.info/index.php/Using_GCCSDK and more
> or less got a working crossbuild environment. I did gave to patch ppl
> for a ptrdiff_t error.

I haven't seen this problem myself; perhaps an issue with the native GCC
you're using to build the cross compiler? I'm using v4.8.2. I've just
recently rebuilt the 4.7.4 compiler from scratch and the only issue was
with downloading the source for GMP which gave this:

--2015-02-26 18:46:54--
ftp://ftp.gmplib.org/pub/gmp-5.0.1/gmp-5.0.1.tar.bz2
=> 'gmp-5.0.1.tar.bz2'
Resolving ftp.gmplib.org (ftp.gmplib.org)... 130.242.124.100
Connecting to ftp.gmplib.org (ftp.gmplib.org)|130.242.124.100|:21...
connected.
Logging in as anonymous ... Logged in!
==> SYST ... done. ==> PWD ... done.
==> TYPE I ... done. ==> CWD (1) /pub/gmp-5.0.1 ... done.
==> SIZE gmp-5.0.1.tar.bz2 ... 2006109
==> PASV ...

It got stuck here, I had to CTRL-C to exit.

Anyone else see this, or is it just me?

I had to change the URL locally in the top level makefile to:

http://ftp.gnu.org/gnu/gmp/gmp-$(GMP_VERSION).tar.bz2

> Attempts then to use the autobuilder bring up
> frequent errors however.
>
> That leads me to ask: is there a more accurate setup script, or has time
> just taken its toll?

Some packages always target the latest source and it's inevitable that
the patches will become outdated causing patching at build time to fail.
OTOH, some packages target a specific version of their source, and the
patches stay stable.

> Also, if patches are found needed, just to build what in theory should
> build, what is the correct feedback path to use?

If any packages that you need fail to build, then it's probably best to
report them here on the mail list and I'm sure someone will have a look
for you.

> A third Q: SharedUnixLibrary: who currently handles this?.. are the
> sources accessible?

This is part of the Unixlib source and can be found at:

gcc4/srcdir/gcc/libunixlib/module/sul.s

Lee.


_______________________________________________
GCCSDK mailing list gcc@gccsdk.riscos.info
Bugzilla: http://www.riscos.info/bugzilla/index.cgi
List Info: http://www.riscos.info/mailman/listinfo/gcc
Main Page: http://www.riscos.info/index.php/GCCSDK

[gccsdk] Building the GCCSDK - current status?

Hi Folks

I thought it was time to look at this. I've followed the various instructions at http://www.riscos.info/index.php/Using_GCCSDK and more or less got a working crossbuild environment. I did gave to patch ppl for a ptrdiff_t error. Attempts then to use the autobuilder bring up frequent errors however.

That leads me to ask: is there a more accurate setup script, or has time just taken its toll? 

Also, if patches are found needed, just to build what in theory should build, what is the correct feedback path to use?

A third Q: SharedUnixLibrary: who currently handles this?.. are the sources accessible?

Many thanks

John


--

Stop press! Just announced -

Free Standard Broadband for 6 months, unlimited data, and no minimum contract

For full details of see http://www.JustAskJB.co.uk

Intrigued? Call me

John Ballance C.Eng MIET - jwb@macpcrepair.co.uk - 07976 295923

Improved cross compilation documentation

We have recently improved the documentation for cross compiling NetSurf.

There is a guide to building NetSurf for RISC OS on Linux here:

http://wiki.netsurf-browser.org/Documentation/BuildingForRISCOSQuickStart

Also, some slightly more general notes have been added in a new cross
compilation section in the QUICK-START guide:

http://source.netsurf-browser.org/netsurf.git/plain/Docs/QUICK-START

--
Michael Drake http://www.netsurf-browser.org/

Re: Malformed site (partly OT).

On Sat, Feb 21, 2015 at 11:47:48PM +0000, Peter Young wrote:
> On 21 Feb 2015 Peter Slegg <p.slegg@scubadivers.co.uk> wrote:
>
> >> Date: Fri, 20 Feb 2015 12:49:38 GMT
> >> From: Peter Young <pnyoung@ormail.co.uk>
> >> Subject: Re: Malformed site (partly OT).
> >>
> >> That's interesting, and indeed the site is misbehaving today. Loading
> >> those three files and then refreshing the site corrects it, so it does
> >> indeed seem that NetSurf is tripping over itself and failing to get
> >> the CSS files.
> >>
> >> Is this worth a bug report?
> >>
> >> Best wishes,
> >>
> >> Peter.
> >>
>
> > I have also seen pages rendered without the css but I sometimes see this
> > on Firefox with my much faster office PC.
>
> So far the site has always worked for me with Windows Chrome. A bug
> report has been sent, and the feedback so far is "Timed out fetching
> the CSS file."

Sorry if this seems a bit short but its really a note for the
developers to remind ourselves what is going on.

The "missing" CSS files in your case are failing to download. The curl
library we use to actually do all the http/https network access is
reporting an error code that tells us the fetch of the CSS file timed
out.

This points to a network issue, now that may be an issue with the RISC
OS network stack interacting with the remote server or simply a
transient error where the connection could not be established.

Debugging networking issues like this is very challenging, especially
on unmaintained niche networking stacks like RISC OS. TCP/IP is an
evolving protocol and sometimes implementations are not completely
backward compatible.

To properly debug this it would be necessary to put a network protocol
analyser on your local network (like wireshark) and perform a full
session analysis which might take a great deal of time to decode.

As NetSurf developers cannot practically do this at best all we can do
is ensure our usage of libcurl is correct and deal with the missing
resources.

The browser itself is correctly handling the missing resource and
showing the page without styling so technically there is no bug within
NetSurf. I am leaving the issue open for now as I think Michael wants
to use it as an excuse to improve our internal error reporting.

>
> Best wishes,
>
> Peter.
>
> --
> Peter Young (zfc Re) and family
> Prestbury, Cheltenham, Glos. GL52, England
> http://pnyoung.orpheusweb.co.uk
> pnyoung@ormail.co.uk
>
>

--
Regards Vincent
http://www.kyllikki.org/

Re: Malformed site (partly OT).

On 21 Feb 2015 Peter Slegg <p.slegg@scubadivers.co.uk> wrote:

>> Date: Fri, 20 Feb 2015 12:49:38 GMT
>> From: Peter Young <pnyoung@ormail.co.uk>
>> Subject: Re: Malformed site (partly OT).
>>
>> That's interesting, and indeed the site is misbehaving today. Loading
>> those three files and then refreshing the site corrects it, so it does
>> indeed seem that NetSurf is tripping over itself and failing to get
>> the CSS files.
>>
>> Is this worth a bug report?
>>
>> Best wishes,
>>
>> Peter.
>>

> I have also seen pages rendered without the css but I sometimes see this
> on Firefox with my much faster office PC.

So far the site has always worked for me with Windows Chrome. A bug
report has been sent, and the feedback so far is "Timed out fetching
the CSS file."

Best wishes,

Peter.

--
Peter Young (zfc Re) and family
Prestbury, Cheltenham, Glos. GL52, England
http://pnyoung.orpheusweb.co.uk
pnyoung@ormail.co.uk

Re: Malformed site (partly OT).

> Date: Fri, 20 Feb 2015 12:49:38 GMT
> From: Peter Young <pnyoung@ormail.co.uk>
> Subject: Re: Malformed site (partly OT).
>
> That's interesting, and indeed the site is misbehaving today. Loading
> those three files and then refreshing the site corrects it, so it does
> indeed seem that NetSurf is tripping over itself and failing to get
> the CSS files.
>
> Is this worth a bug report?
>
> Best wishes,
>
> Peter.
>

I have also seen pages rendered without the css but I sometimes see this
on Firefox with my much faster office PC.

Peter

[gccsdk] [Bug 257] GCC 4.7.4 fails when compling whetstone.c

http://www.riscos.info/bugzilla3/show_bug.cgi?id=257

Lee <leenoar@sky.com> changed:

What |Removed |Added
----------------------------------------------------------------------------
CC| |leenoar@sky.com

--- Comment #1 from Lee <leenoar@sky.com> 2015-02-21 06:36:18 PST ---
This appears to be caused by the call to log() at line 339.

Try compiling with -fno-builtin-log as a work around.

--
Configure bugmail: http://www.riscos.info/bugzilla3/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching all bug changes.

_______________________________________________
GCCSDK mailing list gcc@gccsdk.riscos.info
Bugzilla: http://www.riscos.info/bugzilla/index.cgi
List Info: http://www.riscos.info/mailman/listinfo/gcc
Main Page: http://www.riscos.info/index.php/GCCSDK

Re: Dictionary.com

On 20 Feb, Nick Roberts wrote:

> The website

> http://www.dictionary.com/

> is supposed to have a field where one can type in a word to look up.
> This field is either not present or invisible in NetSurf CI#2600.

It is there, but unlabelled and not at all obvious. It's right under
the word "Dictionary" in the blue banner at the top of the page.

That's with Build 2600 with JS On on RO 5.20 if that's relevant.
--
Brian Howlett
-------------------------------------------------
Watch out...
...you might get what you're after...

Re: Dictionary.com

In article <defcef9854.tigger@bc63.orpheusinternet.co.uk>,
Nick Roberts <tigger@orpheusinternet.co.uk> wrote:
> The website

> http://www.dictionary.com/

> is supposed to have a field where one can type in a word to look up.
> This field is either not present or invisible in NetSurf CI#2600.

This is a display issue; I'm not sure whether CSS or Javascript is the
culprit. In my PC browsers the top, blue bar contains the logo and
"Dictionary", "Thesaurus", "Translator" and "More..." legends all aligned
centrally. Beneath, and still within the top bar is the search field. In
the Netsurf rendition the field is still there, directly beneath those
legends albeit without the magnifying glass icon and displayed as a white
box at the top of the grey left sidebar. Typing a word into that area and
hitting return leads to a new definition page which is also malformed in
the Netsurf view but at least with the definitions displayed.

Also using Netsurf CI#2600

--
_____________________________________________________________________

Brian Jordan
Virtual RPC-AdjustSA on Windows 8.1 Pro
RISC OS 6.20
_____________________________________________________________________

Re: Dictionary.com

On Fri, 20 Feb 2015 14:26:15 GMT, Nick Roberts wrote:

> The website
>
> http://www.dictionary.com/
>
> is supposed to have a field where one can type in a word to look up.
> This field is either not present or invisible in NetSurf CI#2600.

It's the big field underneath "Dictionary". On my screen here I can
see the start of it and then it disappears behind the rest of the
page. If I reduce the width of the window the entire thing disappears
behind - I can still click in and type, but the text does not display.

It's not quite right, but the field is present.

Chris

Re: [gccsdk] Flex and Bison broken?

In message <DUB120-DS2E5C5C93CC121EC758255F02A0@phx.gbl>
alan buckley <alan_baa@hotmail.com> wrote:


>
> I finally got a chance to try a rebuild of m4 and bison and unfortunately
> both need their autobuilder recipes updated for the latest version/compiler
> so won't build.
>

I've had them 'building' OK with the 4.1.2 cross-compiler, cant remember if all the patches lined up or not, sometimes the target area for patches moves on newer versions.

Reportedly a positive for byacc, where it doesn't get locked into a compiler version to work, so it is possible ther may be gcc version dependency.

> Though having re-read the thread it seems your initial problem was with
> Make. Have you solved this with the variables you have found? Or does
> Make need to look for things on RISC OS in a more intelligent way?
>
Dont think there was any make related problems.
Will initially was baffled by the paths that bison was using to find the pkgdatafiles and the m4 binary, but these can be set in the environment.

> Have you some simple Makefile(s) and other source files for the other
> tools so I can try them out and see what is happening myself?
>
Other tools? are you talking about byacc, reflex alternatives.
There is nothing much to see though, they just compile and work.
This is a solution for when bison is used as yacc, but I imagine no good for source code that includes m4 macros that get used.

> I've never used m4, flex or bison in any details, but a few simple
> examples that show what isn't working would help.
>
> Regards,
> Alan

Part of Bison source, there is examples, but any useage will show the problems.
Though the error is not explicit, I believe the issue is with subpipe/exec arrangement not getting the environment or arguments to m4 properly, and I dont think it gets anything sensible back from m4 either.

My approach to this problem is to merge the gnu m4 source into the gnu bison source, changing names in a few cases (symtab.c to m4symtab.c, output.c to m4output.c) and I'm hoping by renaming the m4 main( to m4main( that I can change the use of exec m4 to an internal function.
The two sources are very similar, there is one conflict with rpl_malloc, but it is not needed by the m4 side once the original configuration is done.
So I remarked it out of the combined config.h
Reconfiguring would be a problem.

I have the combined binary/libs building OK, it just needs the work done to call the m4main function through the subpipe/fork arrangement. It will be down to wether fork or vfork can meet the demand I think.

I dont mind being made to look foolish if you can get bison/m4 to work the normal way though (-:

Ron M.



_______________________________________________
GCCSDK mailing list gcc@gccsdk.riscos.info
Bugzilla: http://www.riscos.info/bugzilla/index.cgi
List Info: http://www.riscos.info/mailman/listinfo/gcc
Main Page: http://www.riscos.info/index.php/GCCSDK

Re: Malformed site (partly OT).

On 20 Feb 2015 Tony Moore <old_coaster@yahoo.co.uk> wrote:

> On 20 Feb 2015, Peter Young <pnyoung@ormail.co.uk> wrote:
>> On 20 Feb 2015 Tony Moore <old_coaster@yahoo.co.uk> wrote:

> [snip]

>>> Next time you see the duff screen, try launching the urls above. If
>>> there is no error, or time-out, it means that files are accessible,
>>> but that NetSurf gave up too soon. (Press f8 to see each file's
>>> content.)
>>
>> That's interesting, and indeed the site is misbehaving today. Loading
>> those three files and then refreshing the site corrects it, so it does
>> indeed seem that NetSurf is tripping over itself and failing to get
>> the CSS files.
>>
>> Is this worth a bug report?

> I think so. Using recent NetSurf builds, I've seen the same effect on
> other sites, so this may be an instance of a more general problem.

OK, will do.

Best wishes,

Peter.

--
Peter Young (zfc Re) and family
Prestbury, Cheltenham, Glos. GL52, England
http://pnyoung.orpheusweb.co.uk
pnyoung@ormail.co.uk

[gccsdk] [Bug 257] New: GCC 4.7.4 fails when compling whetstone.c

http://www.riscos.info/bugzilla3/show_bug.cgi?id=257

Summary: GCC 4.7.4 fails when compling whetstone.c
Product: GCC/GCCSDK
Version: other
Platform: Other
OS/Version: RISC OS
Status: NEW
Severity: normal
Priority: P1
Component: C compiler
AssignedTo: John.Tytgat@aaug.net
ReportedBy: beerling@online.de
Estimated Hours: 0.0


The standard whetstone benchmark can be found at
www.netlib.org/benchmark/whetstone.c

when compiling using
gcc -O3 -o whetstone whetstone.c -mlibscl

the compiler hangs, no message nor output is produced.

System:
RISC OS 5.20 (10-Jun-13)
Beagleboard xM
compiling in task window with 30000k memory (works with other programs).

--
Configure bugmail: http://www.riscos.info/bugzilla3/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching all bug changes.

_______________________________________________
GCCSDK mailing list gcc@gccsdk.riscos.info
Bugzilla: http://www.riscos.info/bugzilla/index.cgi
List Info: http://www.riscos.info/mailman/listinfo/gcc
Main Page: http://www.riscos.info/index.php/GCCSDK

Dictionary.com

The website

http://www.dictionary.com/

is supposed to have a field where one can type in a word to look up.
This field is either not present or invisible in NetSurf CI#2600.

--
Nick Roberts tigger @ orpheusinternet.co.uk

Hanlon's Razor: Never attribute to malice that which
can be adequately explained by stupidity.

Re: Malformed site (partly OT).

On 20 Feb 2015, Peter Young <pnyoung@ormail.co.uk> wrote:
> On 20 Feb 2015 Tony Moore <old_coaster@yahoo.co.uk> wrote:

[snip]

> > Next time you see the duff screen, try launching the urls above. If
> > there is no error, or time-out, it means that files are accessible,
> > but that NetSurf gave up too soon. (Press f8 to see each file's
> > content.)
>
> That's interesting, and indeed the site is misbehaving today. Loading
> those three files and then refreshing the site corrects it, so it does
> indeed seem that NetSurf is tripping over itself and failing to get
> the CSS files.
>
> Is this worth a bug report?

I think so. Using recent NetSurf builds, I've seen the same effect on
other sites, so this may be an instance of a more general problem.

Tony

Re: Malformed site (partly OT).

On 20 Feb 2015 Tony Moore <old_coaster@yahoo.co.uk> wrote:

> On 19 Feb 2015, Peter Young <pnyoung@ormail.co.uk> wrote:

> [snip]

>> Either works or doesn't work here for me, in a random fashion. Tony's
>> adjust-click on the reload button doesn't work, sadly.

> I did say 'may':)

Indeed, you did!

> The site's css files are at

> http://www.mssociety.org.uk/sites/default/files/css/css_ae906705ee3398
> da03236704b02fa5cc.css

> http://www.mssociety.org.uk/sites/default/files/css/css_5109959e3208e9
> 6651a9e2f4eb04c72f.css

> http://www.mssociety.org.uk/sites/default/files/css/css_6caea41da8ff09
> b5b9fc4ee406a47fba.css

> Next time you see the duff screen, try launching the urls above. If
> there is no error, or time-out, it means that files are accessible, but
> that NetSurf gave up too soon. (Press f8 to see each file's content.)

That's interesting, and indeed the site is misbehaving today. Loading
those three files and then refreshing the site corrects it, so it does
indeed seem that NetSurf is tripping over itself and failing to get
the CSS files.

Is this worth a bug report?

Best wishes,

Peter.

--
Peter Young (zfc Re) and family
Prestbury, Cheltenham, Glos. GL52, England
http://pnyoung.orpheusweb.co.uk
pnyoung@ormail.co.uk

Re: [Rpcemu] RPCEmu 0.8.9, timers and sound is OK on Win-XP

On 20 Feb 2015, "Timmermans, Andre" <andre.timmermans@atos.net> wrote:

[snip]

> Now if someone could find a fix for RPCEmu being killed when trying to
> go in full screen mode on Windows 7, That would be perfect.

I'm not aware of any fix, but a way round the problem is to use a mode
definition file which causes the RPCEmu window to fill the screen, when
maximised. eg for a Win7 laptop, having a 1280 x 800 screen, I use

# 1280 x 760
startmode
mode_name:1280 x 760 (60.0 Hz)
x_res:1280
y_res:760
pixel_rate:72969
h_timings:188,44,0,1280,0,0
v_timings:2,32,0,760,0,11
sync_pol:3
endmode

The difference of 40 pixels, in vertical resolutions, allows space for
the Windows title bar.

Tony




_______________________________________________
Rpcemu mailing list
Rpcemu@riscos.info
http://www.riscos.info/cgi-bin/mailman/listinfo/rpcemu

Re: [gccsdk] Flex and Bison broken?

> Ron wrote on Monday, February 16, 2015 10:19 PM:

> In message <DUB120-W387DD353C097BCD506734EF0260@phx.gbl>
> alan buckley <alan_baa@hotmail.com> wrote:

> > WPB wrote on Date: Sat, 7 Feb 2015 20:27:31:
> > > Guess what? I'm back!
> > >
> > > So, now with everyone's help (thank you), I've progressed to the point
> > > that 'make' is trying to invoke flex and then bison.
> > >
> > > Both fail. Flex with the following error:
> > >
> > > File '$.usr.local.bin.m4' not found
> > >
> > > I know there is a port of 'm4' at riscos.info, but it's certainly not
> > > going to be found at that location on my hard drive! From my limited
> > > experience of Linux, that looks like a Linux path converted to a RISC
> > > OS
> > > path. I get the same error when trying a very simple invocation of
> > > 'flex'
> > > from the command line. Would people agree that 'flex' is broken, or am
> > > I
> > > doing something wrong?
> > >
> > > So, I plugged in Rik Griffin's port of 'flex'
> > > (http://homepage.ntlworld.com/rik.griffin/) and if I change the paths
> > > appropriately (because it's a Norcroft-compiled version, so has no
> > > sfix
> > > swapping implemented) I can get it to work.
> > >
> > > Then we get to Bison, which fails similarly:
> > >
> > > bison -t -v -b parser -d -o OBJECTS/parser.tab.c src/parser.y
> > > src/parser.y:93.17-24: warning: symbol OPERATOR redeclared
> > > src/parser.y:94.15-23: warning: symbol MACROTEXT redeclared
> > > src/parser.y:263.1: warning: a ';' might be needed at the end of
> > > action
> > > code
> > > src/parser.y:263.1: future versions of Bison will not add the ';'
> > > src/parser.y:275.1: warning: a ';' might be needed at the end of
> > > action
> > > code
> > > src/parser.y:275.1: future versions of Bison will not add the ';'
> > > src/parser.y:522.1: warning: a ';' might be needed at the end of
> > > action
> > > code
> > > src/parser.y:522.1: future versions of Bison will not add the ';'
> > > /HostFS::HostFS.$/AppsInUse/Dev/GCC474r1/!GCC/bin/bison:
> > > /home/alanb/gccsdk/env/share/bison/m4sugar/m4sugar.m4: cannot open: No
> > > such file or directory
> > > make: *** [OBJECTS/parser.tab.c] Error 1
> > >
> > > So it looks like there's a hard-coded reference in there to a file:
> > > /home/alanb/gccsdk/env/share/bison/m4sugar/m4sugar.m4 - again, it
> > > appears
> > > 'bison' is broken. Would people agree?
> > >
> > > If anyone has any input on this, it would be greatly appreciated.
> > >
> >
> > The path looks like it's from my machine when I last ran the autobuilder
> > on bison. If anyone knows what the paths should be for RISC OS (or
> > better still had a patch) I can look at fixing and rebuilding both flex
> > and bison.
> >

> I posted elsewhere that the BISON_PKGDATADIR and M4 environment variables
> can be used.
> However I've found that though m4 will compile and work on it's own
> happily,
> Bison has a complex subpipe system to use m4 as a filter.
> Flex is failing for similar reasons I think.

> The Berkely YACC compiles easily (and works), is available from Debian,
> and reportedly the best YACC available. There are a few 'tricks' that >
> Bison can do, but wether these will be missed remains to be seen, Bison/M4
> is very long winded compared to byacc which is self contained > > and only
> uses temp files. (no pipes, simpler source).

> I found a variant of flex called reflex which seems to be similarly good
> so far.
> The pair have worked with a Bison example, so they look promising.

> I noticed Autoconf uses m4 so it would probably have complications also.

I finally got a chance to try a rebuild of m4 and bison and unfortunately
both need their autobuilder recipes updated for the latest version/compiler
so won't build.

Though having re-read the thread it seems your initial problem was with
Make. Have you solved this with the variables you have found? Or does
Make need to look for things on RISC OS in a more intelligent way?

Have you some simple Makefile(s) and other source files for the other
tools so I can try them out and see what is happening myself?

I've never used m4, flex or bison in any details, but a few simple
examples that show what isn't working would help.

Regards,
Alan


_______________________________________________
GCCSDK mailing list gcc@gccsdk.riscos.info
Bugzilla: http://www.riscos.info/bugzilla/index.cgi
List Info: http://www.riscos.info/mailman/listinfo/gcc
Main Page: http://www.riscos.info/index.php/GCCSDK

Re: Malformed site (partly OT).

On 19 Feb 2015, Peter Young <pnyoung@ormail.co.uk> wrote:

[snip]

> Either works or doesn't work here for me, in a random fashion. Tony's
> adjust-click on the reload button doesn't work, sadly.

I did say 'may':)

The site's css files are at

http://www.mssociety.org.uk/sites/default/files/css/css_ae906705ee3398da03236704b02fa5cc.css

http://www.mssociety.org.uk/sites/default/files/css/css_5109959e3208e96651a9e2f4eb04c72f.css

http://www.mssociety.org.uk/sites/default/files/css/css_6caea41da8ff09b5b9fc4ee406a47fba.css

Next time you see the duff screen, try launching the urls above. If
there is no error, or time-out, it means that files are accessible, but
that NetSurf gave up too soon. (Press f8 to see each file's content.)

Tony

NetSurf 3.3 release

It is six months since the 3.2 release and seems a reasonable time to
release 3.3 which fixes many issues [1]. I will, once again, be the
release manager and intend to perform the release process on Saturday
the 28th February 2015.

The targeted release architectures currently include RISC OS although
I have serious reservations about releasing on this platform. This is
principally because the current maintainer has not had sufficient time
to address the numerous bugs currently assigned [2].

This also raises the question of who (if anyone) is assessing the
impact of core changes on the RISC OS frontend. If the situation does
not improve I shall be forced to remove RISC OS as a release
architecture for subsequent releases.

To be clear, this does not affect it being built as part of the
development process, just having releases provided by NetSurf.

The Win32, Atari and Haiku frontends will not be release
frontends unless a maintainer becomes apparent pretty quickly.

[1] http://bugs.netsurf-browser.org/mantis/changelog_page.php?version_id=15
[2] http://bugs.netsurf-browser.org/mantis/search.php?project_id=0&handler_id=144&sticky_issues=on&sortby=last_updated&dir=DESC&hide_status_id=80&match_type=0
--
Regards Vincent
http://www.kyllikki.org/

Re: [Rpcemu] RPCEmu 0.8.9, timers and sound is OK on Win-XP

Hi,

I have moved to a machine using Windows 7 and progressed to RPCEmu 0.8.12 now but did not make any change.
But today a quick test of DigitalCD on a pre-configured setup with RISC OS 5 IOMD I downloaded a few weeks ago
and noticed that I got no problems. I then started my usual RPCEmu setup, problems.

I have checked the logs of both instance and the problematic one mentioned a bridge.
I then edited the rpc.cfg of my usual setup. Networking was already off but I removed the line with the ipadress
and restarted my usual setup. No more sound problems.

Now if someone could find a fix for RPCEmu being killed when trying to go in full screen mode on Windows 7,
That would be perfect.

Kind Regards,
Timmermans André



-----Original Message-----
From: rpcemu-bounces@riscos.info [mailto:rpcemu-bounces@riscos.info] On Behalf Of Timmermans, Andre
Sent: Thursday, January 12, 2012 11:31 AM
To: rpcemu@riscos.info
Subject: [Rpcemu] RPCEmu 0.8.9, timers and sound is OK on Win-XP

Hi,

This is just a little report concerning the timer/sound problems on Windows XP.

The previous version of RPCEmu had some problems which caused problems like te timers being updated below real time and sound DMA buffers filling not being called often enough, causing stuttering while playing sounds.

I am glad to report that RPCEmu 0.8.9 works fine on my machine.
Since it is not mentionned in the list of changes for this version I can only assume that the fix comes from the newer version of the allegro DLL.

I have noticed some problems with regard to accessing the CD drive of my machine but I will report it later when I had some time to provide some test cases.

Cheers,
André

_______________________________________________
Rpcemu mailing list
Rpcemu@riscos.info
http://www.riscos.info/cgi-bin/mailman/listinfo/rpcemu

This e-mail and the documents attached are confidential and intended solely for the addressee; it may also be privileged. If you receive this e-mail in error, please notify the sender immediately and destroy it. As its integrity cannot be secured on the Internet, Atos' liability cannot be triggered for the message content. Although the sender endeavours to maintain a computer virus-free network, the sender does not warrant that this transmission is virus-free and will not be liable for any damages resulting from any virus transmitted. On all offers and agreements under which Atos Belgium N.V./S.A. supplies goods and/or services of whatever nature, the General Terms and Conditions from Atos Belgium N.V./S.A . exclusively apply. The General Terms and Conditions shall be promptly submitted to you on your request.

Atos Belgium N.V/S.A
Da Vincilaan 5
B-1930 Zaventem
RPR-RPM Brussels
BTW/TVA: BE 0401 848 135
_______________________________________________
Rpcemu mailing list
Rpcemu@riscos.info
http://www.riscos.info/cgi-bin/mailman/listinfo/rpcemu

Re: Malformed site (partly OT).

On 19 Feb 2015 Alan Calder <alan_calder@o2.co.uk> wrote:

> In article <44f5489854.pnyoung@pnyoung.ormail.co.uk>, Peter Young
> <pnyoung@ormail.co.uk> wrote:
>> I maintain the website of the local branch of the Multiple Sclerosis
>> Society, at www.mssociety.org.uk/cheltenham (NB I am only responsible
>> for the content, not the formatting). From time to time the site gets
>> seriously malformed in RISC OS NetSurf, and then a few days later goes
>> back to what is should be. Screenshots of this are at
>> http://pnyoung.orpheusweb.co.uk/Chrome.jpg as it should be in Chrome on
>> Windows and http://pnyoung.orpheusweb.co.uk/NetSurf.jpg as it was
>> yesterday in NetSurf #2600, but it's back to how it should be this
>> morning! Same NetSurf build.

> [Snip]

> Seems to work just fine for me here using NS 3.2, Strongarm RPC with OS
> 4.39. Tried it several times and always seems to come up OK. Not sure
> that all the functionality is there but the bits I tried seemed to work OK.

Either works or doesn't work here for me, in a random fashion. Tony's
adjust-click on the reload button doesn't work, sadly.

Best wishes,

Peter.

--
Peter Young (zfc Re) and family
Prestbury, Cheltenham, Glos. GL52, England
http://pnyoung.orpheusweb.co.uk
pnyoung@ormail.co.uk

Re: Malformed site (partly OT).

On 19 Feb 2015 Tim Hill <tim@timil.com> wrote:

> In article <44f5489854.pnyoung@pnyoung.ormail.co.uk>, Peter Young
> <pnyoung@ormail.co.uk> wrote:
>> I maintain the website of the local branch of the Multiple Sclerosis
>> Society, at www.mssociety.org.uk/cheltenham (NB I am only responsible
>> for the content, not the formatting). From time to time the site gets
>> seriously malformed in RISC OS NetSurf, and then a few days later goes
>> back to what is should be. Screenshots of this are at
>> http://pnyoung.orpheusweb.co.uk/Chrome.jpg as it should be in Chrome on
>> Windows and http://pnyoung.orpheusweb.co.uk/NetSurf.jpg as it was
>> yesterday in NetSurf #2600, but it's back to how it should be this
>> morning! Same NetSurf build.

> That difference is CSS. Perhaps someone knows why NetSurf doesn't always
> seem to see their style sheet(s). I have seen this once or twice, even
> in Windows browsers, when the CSS is missing or misnamed. Perhaps just a
> slow/busy server and NetSurf is impatient?

The CSS do seem to be causing the problem. Does NetSurf always fetch
the components of a site in the same order? Knowledge of these things
is beyond me.


>> I imagine this is the fault of the site rather than of NetSurf, but I
>> would value comments. (OT) If it is the fault of the site, I have a
>> feeling that they may be contravening guidance about website
>> accessibility for disabled people, and I would be interested in
>> comments about this also.

> Their cookie top bars are twee, one is superfluous and both are invisible
> to NetSurf but to answer your question:

The cookie bars are a pain, in that they keep returning every time I
log in on Windows. Surely there should be a cookie to say that I have
agreed. I'll nag the web team about this when I get a round tuit. At
least as you say, I don't see them in NetSurf.

[snip]

Best wishes,

Peter.


--
Peter Young (zfc Re) and family
Prestbury, Cheltenham, Glos. GL52, England
http://pnyoung.orpheusweb.co.uk
pnyoung@ormail.co.uk

RE: Malformed site (partly OT).

In message <f351559854.pnyoung@pnyoung.ormail.co.uk>
Peter Young <pnyoung@ormail.co.uk> wrote:

> On 19 Feb 2015 Dave Higton <dave@davehigton.me.uk> wrote:

>> On Thu, 19 Feb 2015 08:01:50 GMT Peter Young wrote:
>>> I maintain the website of the local branch of the Multiple Sclerosis
>>> Society, at www.mssociety.org.uk/cheltenham (NB I am only responsible
>>> for the content, not the formatting). From time to time the site gets
>>> seriously malformed in RISC OS NetSurf, and then a few days later goes
>>> back to what is should be. Screenshots of this are at
>>> http://pnyoung.orpheusweb.co.uk/Chrome.jpg as it should be in Chrome on
>>> Windows and
>>> http://pnyoung.orpheusweb.co.uk/NetSurf.jpg as it was yesterday in
>>> NetSurf #2600, but it's back to how it should be this morning! Same
>>> NetSurf build.

>> Isn't this what happens when a site doesn't respond quickly enough
>> and Netsurf times out getting later parts of the site?

>> (Not to blame the site for the delay - it could equally be any part
>> of the network between the site and NS.)

> That's an interesting thought, thanks. The site is very slow in
> rendering, even in Chrome; it gets very irritating sometimes, when I'm
> editing it.

Using VRPC on the iMac and NetSurf 2.9, the latest version that works
in that platform, the site renders in 1.3s, and reloads in 0.2s using
the old-fashioned cache. Firefox and Safari are all but instant on the
same iMac.

On the Raspberry Pi with NetSurf 3.3 #2600 the site takes 30s and even
then only renders in the cut down form. NetSurf 3.1, however, on the
Raspberry Pi 2 renders fully in 3.2s.
--
David Pitt
OS4.39, VRPC, iMac

Re: Malformed site (partly OT).

In article <44f5489854.pnyoung@pnyoung.ormail.co.uk>, Peter Young
<pnyoung@ormail.co.uk> wrote:
> I maintain the website of the local branch of the Multiple Sclerosis
> Society, at www.mssociety.org.uk/cheltenham (NB I am only responsible
> for the content, not the formatting). From time to time the site gets
> seriously malformed in RISC OS NetSurf, and then a few days later goes
> back to what is should be. Screenshots of this are at
> http://pnyoung.orpheusweb.co.uk/Chrome.jpg as it should be in Chrome on
> Windows and http://pnyoung.orpheusweb.co.uk/NetSurf.jpg as it was
> yesterday in NetSurf #2600, but it's back to how it should be this
> morning! Same NetSurf build.

[Snip]

Seems to work just fine for me here using NS 3.2, Strongarm RPC with OS
4.39. Tried it several times and always seems to come up OK. Not sure
that all the functionality is there but the bits I tried seemed to work OK.

Alan

--
Alan Calder, Milton Keynes, UK.

Re: Malformed site (partly OT).

In article <44f5489854.pnyoung@pnyoung.ormail.co.uk>, Peter Young
<pnyoung@ormail.co.uk> wrote:
> I maintain the website of the local branch of the Multiple Sclerosis
> Society, at www.mssociety.org.uk/cheltenham (NB I am only responsible
> for the content, not the formatting). From time to time the site gets
> seriously malformed in RISC OS NetSurf, and then a few days later goes
> back to what is should be. Screenshots of this are at
> http://pnyoung.orpheusweb.co.uk/Chrome.jpg as it should be in Chrome on
> Windows and http://pnyoung.orpheusweb.co.uk/NetSurf.jpg as it was
> yesterday in NetSurf #2600, but it's back to how it should be this
> morning! Same NetSurf build.

That difference is CSS. Perhaps someone knows why NetSurf doesn't always
seem to see their style sheet(s). I have seen this once or twice, even
in Windows browsers, when the CSS is missing or misnamed. Perhaps just a
slow/busy server and NetSurf is impatient?

> I imagine this is the fault of the site rather than of NetSurf, but I
> would value comments. (OT) If it is the fault of the site, I have a
> feeling that they may be contravening guidance about website
> accessibility for disabled people, and I would be interested in
> comments about this also.

Their cookie top bars are twee, one is superfluous and both are invisible
to NetSurf but to answer your question:

They seem to meet all their claims to accessibility if I use a Windows
mainstream browser but some of their claims are reliant on a browser's
ability to, say, cycle through all links with the <tab> and select them
with <return>. This is common: is that claim met when NetSurf doesn't
provide that facility? In common with much corporate drivel on
accessibility, I wonder whether their accessibility page should have a
disclaimer about 'depending on browser' and not appear to take credit for
a browser feature! All they have done is get things in the right order.
However, as your second image shows, the basic HTML layout seems tidy and
structured which is pretty much what most accessibility implementation is
really all about. Judging by the inhuman CSS file names used nobody codes
any of their pages except via a CMS.

Interesting that text-low-graphics still imposes their choice of layout
and font on the reader. A designer who won't let go? (S/he won't use
standard G+, facebook, Twitter, or YouTube gfx). I thought that partially
sighted readers were meant to be able to use their browser default font?
To all intents and purposes a vanilla page should be the page with no CSS
and only one menu. Having seen a partially sighted user with a huge
monitor and text blown up to four lines a screenful, I learnt to
appreciate the need for a 'basic' option (appreciate =/= implement
always). Scrolling around a cheesy, classic over-menued 'square' page
(skyscraper menu left, menu bar at the top, menu bar at the bottom with
center and right skyscraper content) when your browser viewport is at
800%+ is not easy as anyone with a small smartphone visiting a
desktop-only web site will attest. Oh, look at sites such as
www.mssociety.org.uk Not really mobile friendly at all. No mobiles at
MSS yet then? ;-)

Try and read that website on a small smartphone and I think anyone would
have comments about practical 'accessibility'.

Where are my specs?

A lookalike MSS website is what you could end up with if you use my ISP's
website builder http://freeola.com/website-builder/ which has lots of
'square' templates with three menus, on one menu thrice.

--

www.timil.com

web sites * multimedia * training