On Tue, Dec 19, 2006 at 03:09:55PM +1800, Bo Peng wrote:
> >> Although your qt installation has to be blamed. I have modified
> >> SConstruct to accommendate this.
> >
> >Why should my qt installation be blamed?
>
> Ok. Maybe it is perfectly ok for a distribution to provide only
> release librarie
Am Dienstag, 19. Dezember 2006 21:39 schrieb Bo Peng:
> > There are indeed only .so versions of iostreams installed (see the list
I
>
> As long as the problem is found, the fix is easy. ... Please check again.
It seems to work now (at least it starts to compile, and prints out that it
found the
There are indeed only .so versions of iostreams installed (see the list I
As long as the problem is found, the fix is easy. ... Please check again.
Bo
Am Dienstag, 19. Dezember 2006 21:00 schrieb Bo Peng:
> It seems to me that scons can not find iostreams... can you check that?
There are indeed only .so versions of iostreams installed (see the list I
sent). I don't know why (the 1.33.1 devel package does not contain a
static library either, s
Checking for boost library signals, regex, filesystem, iostreams... Looking
into $LOCALLIBPATH
Looking into /usr/lib
Find boost libraries:
['/usr/lib/libboost_signals-gcc-1_34.a',
'/usr/lib/libboost_signals-gcc-mt-1_34.a']
Qualified libraries:
['/usr/lib/libboost_signals-gcc-1_34.a',
'/usr/lib
Am Dienstag, 19. Dezember 2006 20:36 schrieb Bo Peng:
> I would appreciate it if you can apply the attached patch and send me
> the result... This boost detection business should not have been this
> hard.
Indeed not.
> Maybe you only have .so files?
No, I also have .a, but it should also work i
Am Dienstag, 19. Dezember 2006 20:29 schrieb Bo Peng:
> The display of (cached) is likely a scons bug. It is displayed because
> my detection does not call any further scons configuration function
> that can override this flag. This is likely the case since scons
> correctly detects boost 1.35 whe
I would appreciate it if you can apply the attached patch and send me
the result... This boost detection business should not have been this
hard.
Maybe you only have .so files?
Thanks.
Bo
Index: development/scons/scons_utils.py
===
I don't really know how you check for qt, but it would be reasonable to
first check whether pkg-config finds qt. Then you don't need to make any
assumptions on linux,
I check pkgconfig for QtCore, and then try to link to QtCore_debug if
mode=debug.
Checking for boost library signals, regex, fi
Am Dienstag, 19. Dezember 2006 19:42 schrieb Bo Peng:
> >
> > No config.log gets created. All I have is this:
>
> This has been fixed. I also did some cleanup of SConstruct along the
> way to have better separation of general (for src etc) and frontend
> (for support and qt4) building environments
No config.log gets created. All I have is this:
This has been fixed. I also did some cleanup of SConstruct along the
way to have better separation of general (for src etc) and frontend
(for support and qt4) building environments.
This problem was not discovered because qt4 has never been insta
Am Dienstag 19 Dezember 2006 16:09 schrieb Bo Peng:
> > > Although your qt installation has to be blamed. I have modified
> > > SConstruct to accommendate this.
> >
> > Why should my qt installation be blamed?
>
> Ok. Maybe it is perfectly ok for a distribution to provide only
> release libraries f
Bo Peng wrote:
> I installed boost 1.33.1 in /usr/lib and scons finds them without
> problem. Maybe your problem is triggered by the coexistence of two
> version of boost?
Well, the build system should be able cope with that.
Georg
I am checking this one by installing system boost 1.33.1. I will get
back to you soon.
I installed boost 1.33.1 in /usr/lib and scons finds them without
problem. Maybe your problem is triggered by the coexistence of two
version of boost?
Bo
> Although your qt installation has to be blamed. I have modified
> SConstruct to accommendate this.
Why should my qt installation be blamed?
Ok. Maybe it is perfectly ok for a distribution to provide only
release libraries for a package like qt-devel. (Then what if people
need to debug qt?) An
Am Montag, 18. Dezember 2006 16:19 schrieb Bo Peng:
> > I still have to comment line 348 in development/scons/qt4.py. If I do
that
> > the linking step succeeds.
>
> Although your qt installation has to be blamed. I have modified
> SConstruct to accommendate this.
Why should my qt installation b
I still have to comment line 348 in development/scons/qt4.py. If I do that
the linking step succeeds.
Although your qt installation has to be blamed. I have modified
SConstruct to accommendate this.
~/lyx-1.5/build-tree-scons$ ./lyx -dbg init
Checking whether LyX is run in place... no
Unable t
Am Samstag, 16. Dezember 2006 20:59 schrieb Bo Peng:
> > Yes, that would be good. BTW the compile worked, but in the final link
> > command it did still use the debug version.
>
> I will explicitly test for debug version before I try to use and test
> them. Hole on.
I still have to comment line 3
Yes, that would be good. BTW the compile worked, but in the final link
command it did still use the debug version.
I will explicitly test for debug version before I try to use and test
them. Hole on.
The version is a snapshot of the 1.34 branch, got it from here:
http://packages.debian.org/exp
Am Samstag, 16. Dezember 2006 20:01 schrieb Bo Peng:
> scons never touch the source directory.
Good to know, I was not sure.
> What linux distribution do you use?
Debian Etch
> I compiled qt 4.12/GPL by myself
> and got the _debug version of the libraries. Of course I can link to
> non-debug
PS: I am deliberately not looking at any scons documentation. If scons
should replace the autotools one day it needs to be as simple to use as a
configure; make; make install; sequence.
Your abuse of scons is very welcome.
Thanks. I want to build like this because I can then use multiple build
Am Samstag, 16. Dezember 2006 17:37 schrieb Bo Peng:
> Your request is heard, and implemented. It turns out to be easy to get
> the location of SConstruct from within scons so you can start scons
> >from any directory now.
Thanks. I want to build like this because I can then use multiple build
d
> Why does it not use the location of SConstruct? I find it quite handy to
> simply call it from the directory where I want to build.
Georg,
Your request is heard, and implemented. It turns out to be easy to get
the location of SConstruct from within scons so you can start scons
from any direct
And how do I specify that I want to build in a different directory? I don't
want to build in the source tree.
Scons *never* touch the source tree. It builds in debug or release
directory for mode=debug (default) or mode=release, and you can use
build_dir=/path/to/build to override the directory
Am Freitag, 15. Dezember 2006 22:29 schrieb Bo Peng:
> In INSTALL.scons, only two ways to fire scons are allowed: from top
> source directory, or from development/scons.
And how do I specify that I want to build in a different directory? I don't
want to build in the source tree.
> This is becau
scons does not work either:
I have updated SConstruct so that scons gives out error messages when
it is not called from top source directory or development/scons.
Please try again and see if you get any other problem.
I am glad that you are trying scons. Just remember that
after the first run,
scons: *** No tool named 'qt4': not a Zip file
Also because scons need to find qt4.py, which is under development/scons.
Bo
scons does not work either:
~/lyx-1.5$ mkdir build-tree-scons
~/lyx-1.5$ cd build-tree-scons
~/lyx-1.5/build-tree-scons$ scons -f
In INSTALL.scons, only two ways to fire scons are allowed: from top
source directory, or from development/scons.
This is because scons tries to locate top source di
On Wed, Dec 13, 2006 at 10:37:08PM +0100, Georg Baum wrote:
> Am Mittwoch, 13. Dezember 2006 18:22 schrieb Bo Peng:
> > > I cannot compile current head (same problems as
> > > Kornel),
> >
> > scons works, so does cmake.
>
> cmake does not work for me because it uses wildcards, and I have files
Am Mittwoch, 13. Dezember 2006 18:22 schrieb Bo Peng:
> > I cannot compile current head (same problems as
> > Kornel),
>
> scons works, so does cmake.
cmake does not work for me because it uses wildcards, and I have files
lying around I don't want to build. scons does not work either:
~/lyx-1.5
Jean-Marc Lasgouttes wrote:
> Did you run autogen.sh?
yes.
Jürgen
Michael Gerz wrote:
> Yes, they also show up on Linux, if you run valgrind. I used valgrind
> last week. Look at Status.15x.
I know (I read your messages), I mean: it does not crash, so there is some OS-
or setup-specifity.
Jürgen
Juergen Spitzmueller schrieb:
Do those three bugs reported by Michael, Edwin and Uwe show some similarities?
Uwe, does this crash also only appear when the table is in a paragraph that
only contains one row? Can you three reproduce the bug of each other?
Yes, they also show up on Linux, if yo
> "Juergen" == Juergen Spitzmueller <[EMAIL PROTECTED]> writes:
Juergen> I cannot compile current head (same problems as Kornel)
Did you run autogen.sh?
JMarc
Georg Baum wrote:
> How did you run configure?
./configure --with-frontend="qt4" --with-qt4-dir=/usr/local/qt4
--with-qt4-includes=/usr/local/qt4/include --disable-pch
--with-version-suffix=-svn --enable-maintainer-mode --disable-stdlib-debug
> Where is your qt installed?
/usr/local/qt4
> I
I cannot compile current head (same problems as
Kornel),
scons works, so does cmake.
Bo
Juergen Spitzmueller wrote:
> Another windows-only crash? I cannot compile current head (same problems
> as Kornel), but I cannot reproduce this with a 3 days old build on Linux.
How did you run configure? Where is your qt installed? I believe I have an
idea, I'll test that later tonight.
Georg
Uwe Stöhr wrote:
> - Create a document with an empty table
> - Mark the whole table
> - insert a minipage in order to have the table inside a minipage
> - crash
Another windows-only crash? I cannot compile current head (same problems as
Kornel), but I cannot reproduce this with a 3 days old build
Juergen Spitzmueller <[EMAIL PROTECTED]> writes:
| Juergen Spitzmueller wrote:
>> Can you please test if the attached patch fixes the things I have promised
>> and if there are no regressions introduced by the patch?
>
| Lars, the patch has been tested by Helge. No regressions have been found. Can
Juergen Spitzmueller wrote:
> Can you please test if the attached patch fixes the things I have promised
> and if there are no regressions introduced by the patch?
Lars, the patch has been tested by Helge. No regressions have been found. Can
I apply it?
Jürgen
Helge Hafting wrote:
> The patch applied to today's cvs (16/3-2005) and compiled fine.
>
> >Helge Hafting wrote:
> >>* select cells, press C-M for math. Crash! No, I don't see how this
> >> could be useful, so ignoring operations on selections when
> >> the current selection doesn't fit the op
Helge Hafting wrote:
> >>* Pressing C-rightarrow (or leftarrow) jumps whole words
> >> at a time, for speedy navigation. Unfortunately the cursor
> >> won't go to the next table cell, I have to release the control key.
> >> It'd be nice to jump to the next cell with C-arrows too.
> >
> >Righ
Helge Hafting wrote:
> Juergen Spitzmueller wrote:
>
>>Right. It's a regression wrt 1.3. Please file a bug report.
>>
>>
> How?
> Went to www.lyx.org, clicked "give feedback", got told
> to please use the mailing lists which I have already done.
Go to http://bugzilla.lyx.org and try to underst
Juergen Spitzmueller wrote:
Right. It's a regression wrt 1.3. Please file a bug report.
How?
Went to www.lyx.org, clicked "give feedback", got told
to please use the mailing lists which I have already done.
I seem to remember some other system for bug reports, but I have
forgotten the details.
Thanks for the reports, Helge. I am answering to both of your mails here,
because I have a patch which fixes some of the issues.
Helge Hafting wrote:
> * select cells, press C-M for math. Crash! ÂNo, I don't see how this
> Â Âcould be useful, so ignoring operations on selections when
> Â Âthe cur
On Fri, 26 Feb 1999, Allan Rae wrote:
> From: Allan Rae <[EMAIL PROTECTED]>
> To: Bruce M Beach <[EMAIL PROTECTED]>
>
> >Hello All
> >
> > I was filling in a table(entering 0,'s and 1's as text)
> > when the pgm dumped core. Of interest is that the emergency
> > save file wa
On Thu, 25 Feb 1999, Bruce M Beach wrote:
>Hello All
>
> I was filling in a table and I thought "What a great idea if .."
> No I'm kidding you.
>
> Start again
> I was filling in a table(entering 0,'s and 1's as text)
> when the pgm dumped core. Of interest is that
Hello All
I was filling in a table and I thought "What a great idea if .."
No I'm kidding you.
Start again
I was filling in a table(entering 0,'s and 1's as text)
when the pgm dumped core. Of interest is that the emergency
save file was truncated to the un
48 matches
Mail list logo
