Jan Hlodan | 1 Apr 12:15 2008
Picon

Re: Minimalistic SPARC KDE 3.5.7 tarball available

Hi Adriaan,

If I hit: startkde, small blue window appears ( it should be KDM ).
That's all.
Nothing works there.
http://pastebin.com/f3b9acaee

Regards,

Jan Hlodan

On Tue, Apr 1, 2008 at 12:27 AM, Adriaan de Groot <groot <at> kde.org> wrote:
> I've just finished using konstruct [1] to build a very minimal KDE 3.5.7 on my
>  U45 (SPARC). The resulting tarball of 95MB is on BionicMutton [4] for the
>  time being. It contains kdelibs, kdebase, kdesdk plus their dependencies --
>  bdb, qt, ... . I have tested this insofar as "PATH=$HOME/kde3.5.7/bin:$PATH ;
>  pkill dtwm ; startkde" actually starts a KDE 3.5.7 session where I can run a
>  konsole. KBabel runs over ssh -X, so that should be fine locally as well.
>
>  In all cases where I ran into annoying dependencies, I have simply removed
>  them. kdepim is somehow a dependency for sdk -- bugger that for a lark.
>
>  Fairly minimal changes were needed to konstruct itself. Patch [2] and BDB GAR
>  [3] are also available. In some cases the compile failed and I had to cd in
>  to fix things -- #include <errno.h> was missing somewhere, for instance. BDB
>  needs an extra cd build_unix before make that I can't finagle into GAR. There
>  might have been some linking or compile problems in KDE code which I don't
>  remember. Generally if a directory is problematic, remove it from the
>  Makefile one level above (or run make -k once, for pod2man issues in kdesdk).
>  Ah, BDB in particular needed some fiddling in order to link to libnsl --
(Continue reading)

Jan Hlodan | 12 Apr 15:14 2008
Picon

Re: How to build a stable version of KDE

Installation Stable Version - KDE 3.5.9

Tested on snv_79b x86 (32bit and 64bit)

export PATH=/usr/perl5/5.8.4/bin/:/usr/local/bin:/usr/sfw/bin/:/usr/openwin/bin/:/bin/:/usr/X11/bin/:/usr/X/bin/:/usr/ucb/:/usr/gnu/bin/:/usr/ccs/bin/:/sbin/:/usr/local/kde/bin:/usr/local/qt/bin:/usr/sbin:/usr/bin:$PATH

export LD_LIBRARY_PATH=/usr/lib:/usr/local/lib:/usr/X/lib/:/usr/X11/lib/:/usr/ccs/lib/:/usr/openwin/lib/:/usr/openwin/sfw/lib/:/usr/sfw/lib/:/lib/:/usr/local/kde/lib:/usr/local/qt/lib:$LD_LIBRARY_PATH

CC=/usr/sfw/bin/gcc
CXX=/usr/sfw/bin/g++
export CXX CC

QT

Download: qt-x11-free-3.3.8

ftp://ftp.trolltech.com/qt/source/

gunzip2 qt-x11-free-3.3.8.tar.bz2
tar xvf qt-x11-free-3.3.8.tar
mv qt-x11-free-3.3.8 qt
mkdir /usr/local
mv qt /usr/local/
cd /usr/local/qt

./configure -system-zlib -qt-gif -system-libpng \
-platform solaris-g++ -prefix /usr/local \
-system-libjpeg -plugin-imgfmt-mng -thread -no-stl \
-no-xinerama -no-g++-exceptions

(Continue reading)

Sebastian Kuegler | 29 Apr 16:25 2008
Picon

Re: Linkage errors in kdebase with xrender

On Saturday 26 April 2008 15:46:43 Allen Winter wrote:
> On Friday 25 April 2008 17:19:34 Shai Berger wrote:
> > This is my first message here. I'm an experienced C++ programmer, have
> > been using Debian (sid) and KDE at home for several years now, and
> > finally I'm starting to get into KDE development. My main interests are
> > kdepim (as a devout user) and BiDi issues (as a Hebrew user).
>
> Thanks Shai,
>
> Committed... but...
> we have a problem in dolphin (dolphin/src/ktooltip.cpp) and plasma
> (plasma/plasma/plasmaapp.cpp) because both of those unconditionally
> assume XRender for any system with X11.  
>
> We need the authors of the relevant code to provide #if (HAVE_XRENDER)
> conditional compiles to handle the situation where the XRender stuff
> isn't available.

Apparently this is also a runtime problem. I've seen the Oxygen style and 
Plasma look funny on a SUNray box, where apparently XRender is not available 
but (tried to be) used, resulting in LSD-like coloring of lots of widgets. 
Not that it doesn't look cool, but it does not a match with the artistic 
vision either. :-)

CC:ing kde-solaris.
--

-- 
sebas

 http://www.kde.org | http://vizZzion.org |  GPG Key ID: 9119 0EF9 
(Continue reading)

Kunal Deo | 30 Apr 07:18 2008
Picon

Compilation problems with Dbus-glib

I am trying to compile dbus-glib package. I am getting following error

elf_begin: I/O error: region read: Is a directory
ld: fatal: File processing errors. No output written to 
.libs/dbus-binding-tool
collect2: ld returned 1 exit status

Please help.
--

-- 
Regards

Kunal

___________________________________________________
This message is from the kde-solaris mailing list.
Account management:  https://mail.kde.org/mailman/listinfo/kde-solaris.
Archives: http://lists.kde.org/.
More info: http://www.kde.org/faq.html.

Kunal Deo | 30 Apr 07:23 2008
Picon

Re: Compilation problems with Dbus-glib

exact compiler output

gmake[4]: Entering directory 
`/home/kd223417/Documents/source/kde4/preq/dbus-glib-0.74/dbus'
/bin/bash ../libtool --tag=CC   --mode=link gcc  -mcpu=v9 -Wall 
-Wchar-subscripts -Wmissing-declarations -Wmissing-prototypes 
-Wnested-externs -Wpointer-arith -Wcast-align -Wfloat-equal 
-Wsign-compare   -o dbus-binding-tool dbus-binding-tool-glib.o 
dbus-glib-tool.o libdbus-gtool.la libdbus-glib-1.la 
/home/kd223417/Documents/source/kde4/ 
-L/home/kd223417/Documents/source/kde4//lib -lgobject-2.0 -lglib-2.0   
-lexpat -lnsl -lsocket
gcc -mcpu=v9 -Wall -Wchar-subscripts -Wmissing-declarations 
-Wmissing-prototypes -Wnested-externs -Wpointer-arith -Wcast-align 
-Wfloat-equal -Wsign-compare -o .libs/dbus-binding-tool 
dbus-binding-tool-glib.o dbus-glib-tool.o 
/home/kd223417/Documents/source/kde4/  ./.libs/libdbus-gtool.a 
./.libs/libdbus-glib-1.so -L/home/kd223417/Documents/source/kde4//lib 
/home/kd223417/Documents/source/kde4//lib/libgobject-2.0.so 
/home/kd223417/Documents/source/kde4//lib/libglib-2.0.so -lexpat -lnsl 
-lsocket -R/home/kd223417/Documents/source/kde4//lib
ld: elf error: file /home/kd223417/Documents/source/kde4/: elf_begin: 
I/O error: region read: Is a directory
ld: fatal: File processing errors. No output written to 
.libs/dbus-binding-tool
collect2: ld returned 1 exit status

Kunal Deo wrote:
> I am trying to compile dbus-glib package. I am getting following error
>
(Continue reading)

krishna | 30 Apr 12:10 2008

automated response

Hello

I am away from 1st May to 4th May 2008. 

During this period, I may not be able to reply your email. 

If the matter on hand requires urgent attention, please contact my colleague, Mr Vijaya Kumar, at
vijayakumar.c <at> zagro.com .
Alternatively you may also call at +65 6759 1811

Please do not send big attachment files during this period..

Regards
Krishna
System Administrator & DBA
Zagro Singapore Pte Ltd
___________________________________________________
This message is from the kde-solaris mailing list.
Account management:  https://mail.kde.org/mailman/listinfo/kde-solaris.
Archives: http://lists.kde.org/.
More info: http://www.kde.org/faq.html.

Matthew Woehlke | 1 May 02:30 2008
Picon
Picon

oxygen style needs xrender? (was: Linkage errors in kdebase with xrender)

Sebastian Kuegler wrote:
> Apparently this is also a runtime problem. I've seen the Oxygen style and 
> Plasma look funny on a SUNray box, where apparently XRender is not available 
> but (tried to be) used, resulting in LSD-like coloring of lots of widgets. 
> Not that it doesn't look cool, but it does not a match with the artistic 
> vision either. :-)

Oxygen style? I didn't know that used xrender? Got a screenshot?

--

-- 
Matthew
ELANG: input is in wrong language (please use English on English lists)

 
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<

Girish Ramakrishnan | 1 May 11:26 2008
Picon

Re: oxygen style needs xrender?

Matthew Woehlke wrote:
> Sebastian Kuegler wrote:
>> Apparently this is also a runtime problem. I've seen the Oxygen style and 
>> Plasma look funny on a SUNray box, where apparently XRender is not available 
>> but (tried to be) used, resulting in LSD-like coloring of lots of widgets. 
>> Not that it doesn't look cool, but it does not a match with the artistic 
>> vision either. :-)
> 
> Oxygen style? I didn't know that used xrender? Got a screenshot?
> 

I believe this is expected behavior. Without xrender, Qt does not 
support antialiasing and some composition modes. When no xrender is 
detected, Qt just falls back to the Windows style. I guess kde needs to 
do similar?

You can test this on any machine using QT_X11_NO_XRENDER=1 ./myapp.

Girish

>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<

David Johnson | 1 May 18:10 2008

Re: oxygen style needs xrender?

On Thursday 01 May 2008 02:26:42 am Girish Ramakrishnan wrote:
> I believe this is expected behavior. Without xrender, Qt does not
> support antialiasing and some composition modes. When no xrender is
> detected, Qt just falls back to the Windows style. I guess kde needs to
> do similar?

Except when the user has explicitly specified a style (in code, with qtconfig, 
or via -style). 

--

-- 
David Johnson

>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<

Matthew Woehlke | 1 May 19:09 2008
Picon
Picon

Re: oxygen style needs xrender?

Girish Ramakrishnan wrote:
> Matthew Woehlke wrote:
>> Sebastian Kuegler wrote:
>>> Apparently this is also a runtime problem. I've seen the Oxygen style and 
>>> Plasma look funny on a SUNray box, where apparently XRender is not available 
>>> but (tried to be) used, resulting in LSD-like coloring of lots of widgets. 
>>> Not that it doesn't look cool, but it does not a match with the artistic 
>>> vision either. :-)
>> Oxygen style? I didn't know that used xrender? Got a screenshot?
> 
> I believe this is expected behavior. Without xrender, Qt does not 
> support antialiasing and some composition modes. 

Ah, ok, so it's indirect via using the nice painting features of Qt...

> When no xrender is 
> detected, Qt just falls back to the Windows style. I guess kde needs to 
> do similar?
> 
> You can test this on any machine using QT_X11_NO_XRENDER=1 ./myapp.

AAIIIEE!!! Ok, that looks HORRIBLE! So, yeah, I guess when picking the 
default style we need to check for XRender and use some generic style if 
it isn't present.

Is the case of "occasionally" having/not having XRender rare enough that 
we can allow a user-selected style (via the KCS) to override that 
behavior? Or should we always check unless -style was used?

--

-- 
(Continue reading)


Gmane