Bug#720569: scala upload
Hi tony, On Thursday 26 December 2013 22:05:03 tony mancill wrote: I just realized that scala is not a pkg-java/Debian Java Team maintained package, and so the upload I just made to get scala to build against Java7 could very rightly be considered rude by the team. (It also means that I can't push the commits to the pkg-scala packaging repo.) the Scala team has been inactive for a very long time, so I guess nobody considers your upload rude. On the contrary, your upload brought Scala back to testing, so thank you for that! I can format-patch the changes to you if you'd like, but there's nothing very complicated going on. My apologies if this disrupts any packaging work that was already in progress. No problem, I just pushed your changes to the Git repo. Thanks again for your help! Frank signature.asc Description: This is a digitally signed message part.
Bug#666337: euler: FTBFS: gcc: error: unrecognized option '-V'
On Friday 01 June 2012 15:23:27 gregor herrmann wrote: On Fri, 01 Jun 2012 14:19:42 +0200, Sebastian Ramacher wrote: If you have the time to create a debdiff, great. Otherwise I can also do it myself. Here it is. LDFLAGS is exported now as well. Otherwise the maintainer's change from 1.61.0-5 is reverted, i.e the unneeded dependencies are back. Good catch. Sorry, that was the wrong patch. Here is the right one. Thank you, uploaded to DELAYED/2. Thank you guys for fixing this bug! signature.asc Description: This is a digitally signed message part.
Bug#527708: wxmaxima: FTBFS: wxchar.h:894: error: expected initializer before 'ATTRIBUTE_PRINTF_3'
On Saturday 27 June 2009 14:14:19 Filippo Giunchedi wrote: On Thu, Jun 25, 2009 at 06:00:38PM +0200, Frank S. Thomas wrote: On Friday 08 May 2009 19:52:37 Lucas Nussbaum wrote: During a rebuild of all packages in sid, your package failed to build on amd64. Relevant part: g++ -DHAVE_CONFIG_H -I. -I. -I. [...] MathParser.cpp:220: error: 'Format' is not a member of 'wxString' make[2]: *** [MathParser.o] Error 1 I'm currently preparing a new upstream release of wxMaxima (0.8.2) which does not exhibit this problem and therefore I'm tagging this bug as pending. it does compile despite http://bugs.debian.org/521924 ? The new upstream version requires wxWidgets = 2.8.0. Cheers Frank signature.asc Description: This is a digitally signed message part.
Bug#527708: wxmaxima: FTBFS: wxchar.h:894: error: expected initializer before 'ATTRIBUTE_PRINTF_3'
package wxmaxima tags 527708 + pending thanks Hi, On Friday 08 May 2009 19:52:37 Lucas Nussbaum wrote: During a rebuild of all packages in sid, your package failed to build on amd64. Relevant part: g++ -DHAVE_CONFIG_H -I. -I. -I. [...] MathParser.cpp:220: error: 'Format' is not a member of 'wxString' make[2]: *** [MathParser.o] Error 1 I'm currently preparing a new upstream release of wxMaxima (0.8.2) which does not exhibit this problem and therefore I'm tagging this bug as pending. Cheers Frank -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#511521: boinc: Does not check the RSA_public_decrypt() return value.
On Tuesday 13 January 2009 16:43:55 Rene Mayorga wrote: I was about to port this fix to 6.2.14-2, are everybody ok with this? FWIW: I'm ok with backporting this to the current version in Lenny. signature.asc Description: This is a digitally signed message part.
Bug#504467: init script fails to shut down boinc if $BOINC_OPTS contains --daemon
reassign 504467 boinc-client 6.2.14-2 severity 504467 normal thanks The boinc-client init script fails to shut down boinc if /var/run/boinc.pid does not contain the actual boinc pid. This can happen if boinc is started with the --daemon option. Since this option is not used in the default configuration of boinc-client, this bug is at most of severity normal and not release critical. An easy fix would be to add a warning to /etc/default/boinc-client to not use the --daemon option. More elegant would be of course if the init script would work with or without this option. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#491752: kde4: No Desktop
On Monday 21 July 2008 22:58, miguel wrote: I've been following kde4-experimental for at least a month now and have had no issues with upgrades until the last one about a week ago. Desktop does not start. After logging in, I have a chequered gray/white canvas for a desktop, but nothing else. I had the same issue here. The problem was that kdelibs5/4:4.0.84-1, which was the last version that was uploaded to experimental, was not upgraded to the latest version in unstable (currently 4:4.0.98+svn833207-1). I did this then manually with apt-get install -t unstable kdelibs5 and all is back to normal now. This also fixed the Could not find the Okular component.-issue I had with okular. Cheers, Frank -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#473965: [Pkg-scicomp-devel] Bug#473965: orsa: FTBFS: libginac.so: undefined reference to `cln::cl_double_to_DF_pointer(cln::dfloatjanus const)'
severity 473965 important block 473965 by 473494 stop On Wednesday 02 April 2008 12:44, Lucas Nussbaum wrote: During a rebuild of all packages in sid, your package failed to build on i386. Relevant part: /usr/lib/libginac.so: undefined reference to `cln::cl_double_to_DF_pointer(cln::dfloatjanus const)' collect2: ld returned 1 exit status ORSA FTBFS because of http://bugs.debian.org/473494. Once #473494 is fixed, this bug can be closed. Cheers, -- Frank S. Thomas [EMAIL PROTECTED] PGP public key ID: 0xDC426429 Debian Developerfinger fst/[EMAIL PROTECTED] signature.asc Description: This is a digitally signed message part.
Bug#470246: labplot: FTBFS: ImportOPJ.cc:47: error: no matching function for call to 'Spreadsheet::setColumnType(int, ColumnType)'
tags 470246 + confirmed patch pending
stop
On Monday 10 March 2008 09:48, Lucas Nussbaum wrote:
During a rebuild of all packages in sid, your package failed to build on
i386.
Relevant part:
ImportOPJ.cc: In member function 'int
ImportOPJ::import()':
ImportOPJ.cc:47: error: no matching function for call to
'Spreadsheet::setColumnType(int, ColumnType)' Spreadsheet.h:69: note:
candidates are: void Spreadsheet::setColumnType(int, QString)
ImportOPJ.cc:53: error: cannot convert 'ColumnType' to 'const char*' for
argument '1' to 'int strcmp(const char*, const char*)' ImportOPJ.cc:65:
error: 'class OPJFile' has no member named 'matrixParentFolder'
ImportOPJ.cc:102: error: 'class OPJFile' has no member named
'noteParentFolder' ImportOPJ.cc:118: error: 'class OPJFile' has no member
named 'graphParentFolder' ImportOPJ.cc:142: error: 'xlabel' was not
declared in this scope ImportOPJ.cc:142: error: expected type-specifier
before 'Label'
ImportOPJ.cc:142: error: expected `;' before 'Label'
ImportOPJ.cc:143: error: 'ylabel' was not declared in this scope
ImportOPJ.cc:143: error: expected type-specifier before 'Label'
ImportOPJ.cc:143: error: expected `;' before 'Label'
ImportOPJ.cc:345: error: conversion from 'graphLayerRange' to non-scalar
type 'std::vectordouble, std::allocatordouble ' requested
ImportOPJ.cc:346: error: conversion from 'graphLayerRange' to non-scalar
type 'std::vectordouble, std::allocatordouble ' requested make[3]:
*** [ImportOPJ.lo] Error 1
The full build log is available from:
http://people.debian.org/~lucas/logs/2008/03/08
This FTBFS is caused by the new liborigin/20080225-1 which changed it's API
and ABI. For example the declaration of the OPJFile::colType() function
changed from
const char *OPJFile::colType(int, int) const; (20071119)
to
ColumnType OPJFile::colType(int, int) const;(20080225).
And this is not the only API/ABI incompatible change. I guess liborigin's
SONAME should have been bumped for version 20080225. However, I've a patch
ready to fix this bug (it is attached, and requires a B-D on liborigin (=
20080225)) but it needs some more testing before I can upload a fixed labplot
package.
Cheers,
--
Frank S. Thomas [EMAIL PROTECTED] PGP public key ID: 0xDC426429
Debian Developerfinger fst/[EMAIL PROTECTED]
diff --git a/src/ImportOPJ.cc b/src/ImportOPJ.cc
index 2b625d4..790d09c 100644
--- a/src/ImportOPJ.cc
+++ b/src/ImportOPJ.cc
@@ -12,6 +12,36 @@
#include liborigin/OPJFile.h
+QString colTypeToString(const ColumnType type) {
+QString type_str = ;
+
+switch (type) {
+case X:
+type_str = X;
+break;
+case Y:
+type_str = Y;
+break;
+case Z:
+type_str = Z;
+break;
+case XErr:
+type_str = DX;
+break;
+case YErr:
+type_str = DY;
+break;
+case Label:
+type_str = LABEL;
+break;
+case NONE:
+type_str = NONE;
+break;
+}
+
+return type_str;
+}
+
ImportOPJ::ImportOPJ(MainWin *mw, QString filename)
: mw(mw),filename(filename)
{}
@@ -44,13 +74,13 @@ int ImportOPJ::import() {
for (int j=0;jnr_cols;j++) {
QString name(opj.colName(s,j));
spread-setColumnTitle(j,name.replace(QRegExp(.*_),));
- spread-setColumnType(j,opj.colType(s,j));
+ spread-setColumnType(j,colTypeToString(opj.colType(s,j)));
for (int i=0;iopj.numRows(s,j);i++) {
double *v = (double *) opj.oData(s,j,i,true);
LTableItem *item;
-if(strcmp(opj.colType(s,j),LABEL)) { // number
+if(strcmp(colTypeToString(opj.colType(s,j)),LABEL)) { // number
if(fabs(*v)0 fabs(*v)2.0e-300) // empty entry
continue;
item = new LTableItem( table, QTableItem::OnTyping,QString::number(*v));
@@ -62,7 +92,7 @@ int ImportOPJ::import() {
}
}
for (int s=0;sopj.numMatrices();s++) {
- kdDebug() Matrix s+1 : opj.matrixName(s) (ParentFolder : opj.matrixParentFolder(s))endl;
+ kdDebug() Matrix s+1 : opj.matrixName(s)endl; // (ParentFolder : opj.matrixParentFolder(s))endl;
kdDebug() Label : opj.matrixLabel(s) Cols/Rows : opj.numMatrixCols(s)'/'opj.numMatrixRows(s)endl;
kdDebug() Formula : opj.matrixFormula(s) DisplayType : opj.matrixNumDisplayType(s)endl;
@@ -99,7 +129,7 @@ int ImportOPJ::import() {
QString notes = mw-getProject()-Notes();
for (int s=0;sopj.numNotes();s++) {
- kdDebug() Note s+1 : opj.noteName(s) (ParentFolder : opj.noteParentFolder(s))endl;
+ kdDebug() Note s+1 : opj.noteName(s)endl; // (ParentFolder : opj.noteParentFolder(s))endl;
kdDebug() Label : opj.noteLabel(s) Text : opj.noteText(s)endl;
notes.append(QString(opj.noteLabel(s))+:\n);
notes.append(opj.noteText(s));
@@ -115,7 +145,7 @@ int ImportOPJ::import() {
}
for (int s=0;sopj.numGraphs();s++) {
- kdDebug() Graph s+1 : opj.graphName(s) (ParentFolder : opj.graphParentFolder(s))endl;
+ kdDebug() Graph s+1
Bug#430750: labplot_1.5.1.6-1: FTBFS: autoconf: command not found
Hi Helen, On Sunday 23 December 2007 10:15:56 Frank S. Thomas wrote: On Wednesday 27 June 2007 05:38:33 Steve Langasek wrote: labplot is failing to build on some architectures with the following error: [...] cd . /usr/bin/make -f admin/Makefile.common configure make[2]: Entering directory `/build/buildd/labplot-1.5.1.6' ./admin/cvs.sh: line 11: autoconf: command not found make[2]: *** [configure] Error 1 [...] As it can be seen in build logs for 1.5.1.6-2 [1] and 1.5.1.6-2.1 [2], this bug hasn't been fixed in 1.5.1.6-2, therefore I'm reopening it and removing the record that it has been fixed in 1.5.1.6-2. I'm currently working on another non-maintainer upload which will address this issue. My second NMU of labplot is now ready. To fix this bug I've changed some things which results in a very large diff to the previous revision (my first NMU, 1.5.1.6-2.1) but has the advantage of a clean .diff.gz which only contains the debian/ directory. The large diff to the previous revision is because I've removed all the remade GNU Build System files from the .diff.gz. The required autoreconfing now happens at build time in the config.status target by running $(MAKE) -f admin/Makefile.common, this seems to be much less error prone than adding the remade build system files directly to the .diff.gz. To avoid that changes of the remade build system files end up in the .diff.gz, I've also removed the config.status target as prerequisite from the clean target. My NMU packages are available at: http://people.debian.org/~fst/NMUs/labplot/ Here is the complete diff to the previous version: http://people.debian.org/~fst/NMUs/labplot_1.5.1.6-2.1_1.5.1.6-2.2.diff This is the same as above but only including the relevant changes in the debian/ directory (this patch also attached to this mail): http://people.debian.org/~fst/NMUs/labplot_1.5.1.6-2.1_1.5.1.6-2.2_debian.diff Merry Christmas! Frank --- labplot-1.5.1.6/debian/changelog +++ labplot-1.5.1.6/debian/changelog @@ -1,3 +1,25 @@ +labplot (1.5.1.6-2.2) unstable; urgency=low + + * Non-maintainer upload (encouraged by maintainer). + * Modified build system to fix a bug caused by patches and autogenerated +files (Closes: #430750). The following has been changed in debian/rules: +- Run $(MAKE) -f admin/Makefile.common in the config.status target to + remake the GNU Build System files at build time as suggested in #415656 + instead of adding the remade files directly in the .diff.gz. This + requires build dependencies on autoconf (= 2.53) and automake1.9. +- Removed config.status as prerequisite from the clean target, so that + changes to the build system files do not end up in the .diff.gz. +- Do not copy config.guess and config.sub into LabPlot's top-level source + directory, it is unnecessary. These are also not in the .orig.tar.gz. + * src/x-lpl.desktop: Reverted the change made in 1.5.1.6-2 because the keys +Patterns and DefaultApp are required for KDE 3's MIME type handling, see +http://techbase.kde.org/Development/Architecture/KDE3/MIME_Types. And also +newer Lintian versions do not warn about these keys as unknown anymore. + * Added debian/watch file using http://sf.net/labplot/ as alias for the +qa.debian.org redirector as documented in uscan's manual page. + + -- Frank S. Thomas [EMAIL PROTECTED] Mon, 24 Dec 2007 13:55:05 +0100 + labplot (1.5.1.6-2.1) unstable; urgency=low * Non-maintainer upload (encouraged by maintainer). --- labplot-1.5.1.6/debian/control +++ labplot-1.5.1.6/debian/control @@ -2,7 +2,7 @@ Section: kde Priority: optional Maintainer: Helen Faulkner [EMAIL PROTECTED] -Build-Depends: debhelper (= 5.0.51), kdelibs4-dev, libfftw3-dev, libgsl0-dev, libtiff4-dev, libmagick9-dev, libmagick++9-dev, libbz2-dev, pstoedit, libpstoedit-dev, netcdfg-dev, libaudiofile-dev, pkg-config, quilt, autotools-dev, libtool +Build-Depends: debhelper (= 5.0.51), kdelibs4-dev, libfftw3-dev, libgsl0-dev, libtiff4-dev, libmagick9-dev, libmagick++9-dev, libbz2-dev, pstoedit, libpstoedit-dev, netcdfg-dev, libaudiofile-dev, pkg-config, quilt, autotools-dev, libtool, autoconf (= 2.53), automake1.9 Build-Conflicts: labplot Standards-Version: 3.7.3 Homepage: http://labplot.sourceforge.net/ --- labplot-1.5.1.6/debian/rules +++ labplot-1.5.1.6/debian/rules @@ -48,9 +48,8 @@ cp -f /usr/share/libtool/ltmain.sh admin/ltmain.sh cp -f /usr/share/misc/config.guess admin/config.guess cp -f /usr/share/misc/config.sub admin/config.sub - # I am not 100 % sure if adding those two lines is really necessary. --Charles - cp -f /usr/share/misc/config.guess config.guess - cp -f /usr/share/misc/config.sub config.sub + + $(MAKE) -f admin/Makefile.common VTK_DIR=/usr ./configure --disable-cdf $(configkde) @@ -68,7 +67,7 @@ touch build-stamp -clean: config.status clean-patched unpatch +clean: clean-patched unpatch clean-patched: dh_testdir dh_testroot @@ -81,13
Bug#430750: labplot_1.5.1.6-1: FTBFS: autoconf: command not found
reopen 430750 ! notfixed 430750 1.5.1.6-2 tags 430750 - patch stop On Wednesday 27 June 2007 05:38:33 Steve Langasek wrote: Package: labplot Version: 1.5.1.6-1 Severity: serious labplot is failing to build on some architectures with the following error: [...] cd . /usr/bin/make -f admin/Makefile.common configure make[2]: Entering directory `/build/buildd/labplot-1.5.1.6' ./admin/cvs.sh: line 11: autoconf: command not found make[2]: *** [configure] Error 1 [...] Hi, As it can be seen in build logs for 1.5.1.6-2 [1] and 1.5.1.6-2.1 [2], this bug hasn't been fixed in 1.5.1.6-2, therefore I'm reopening it and removing the record that it has been fixed in 1.5.1.6-2. I'm currently working on another non-maintainer upload which will address this issue. Grüße, Frank [1,2]: http://buildd.debian.org/fetch.cgi?pkg=labplotver=1.5.1.6-2arch=amd64stamp=1188787078file=log http://buildd.debian.org/fetch.cgi?pkg=labplotver=1.5.1.6-2.1arch=powerpcstamp=1198380765file=log signature.asc Description: This is a digitally signed message part.
Bug#431522: depends on non-essential package ucf in postrm
Hello Paul, I have prepared a non-maintainer upload to fix this bug (#431522) and an other serious issue: po-debconf is listed in Build-Depends-Indep instead of Build-Depends although it is required for debian/rules clean. Debian Policy 7.6 requires that dependencies that are required for the clean target must be listed in Build-Depends. My NMU packages and the diff to your last upload (which is also attached to this mail) is available at: http://people.debian.org/~fst/NMUs/localepurge/ http://people.debian.org/~fst/NMUs/localepurge_0.5.9_0.5.9-0.1.diff Although we are currently in an everlasting BSP with a 0-day NMU policy (see [1]), I have uploaded my NMU to DELAYED/7-day for your convenience if you want to cancel my NMU. I hope this in your interest and you do not feel threatened by my NMU. Greetings, Frank [1] http://lists.debian.org/debian-devel-announce/2007/09/msg0.html -- Die Garde stirbt, aber sie ergibt sich nicht! diff -Nru /tmp/gDkhrPCi1T/localepurge-0.5.9/debian/changelog /tmp/o1K6XbbjEe/localepurge-0.5.9/debian/changelog --- /tmp/gDkhrPCi1T/localepurge-0.5.9/debian/changelog 2007-03-19 11:28:38.0 +0100 +++ /tmp/o1K6XbbjEe/localepurge-0.5.9/debian/changelog 2007-12-10 00:12:02.0 +0100 @@ -1,3 +1,16 @@ +localepurge (0.5.9-0.1) unstable; urgency=medium + + * Non-maintainer upload to fix RC bugs. + + * Mask the ucf call in debian/postrm so that the package does not fail on +purge if ucf is not available. (closes: #431522) + + * Moved po-debconf from Build-Depends-Indep to Build-Depends because it is +required to run the debian/rules clean target and dependencies required +for this target must be listed in Build-Depends, see Debian Policy 7.6. + + -- Frank S. Thomas [EMAIL PROTECTED] Sun, 09 Dec 2007 23:34:56 +0100 + localepurge (0.5.9) unstable; urgency=high * First release after moving from Germany to Barcelona/Spain. :-) diff -Nru /tmp/gDkhrPCi1T/localepurge-0.5.9/debian/control /tmp/o1K6XbbjEe/localepurge-0.5.9/debian/control --- /tmp/gDkhrPCi1T/localepurge-0.5.9/debian/control 2006-10-20 23:32:54.0 +0200 +++ /tmp/o1K6XbbjEe/localepurge-0.5.9/debian/control 2007-12-10 00:12:02.0 +0100 @@ -2,7 +2,8 @@ Section: admin Priority: optional Maintainer: Paul Seelig [EMAIL PROTECTED] -Build-Depends-Indep: debmake, po-debconf +Build-Depends: po-debconf +Build-Depends-Indep: debmake Standards-Version: 3.5.10 Package: localepurge diff -Nru /tmp/gDkhrPCi1T/localepurge-0.5.9/debian/postrm /tmp/o1K6XbbjEe/localepurge-0.5.9/debian/postrm --- /tmp/gDkhrPCi1T/localepurge-0.5.9/debian/postrm 2006-10-20 23:32:54.0 +0200 +++ /tmp/o1K6XbbjEe/localepurge-0.5.9/debian/postrm 2007-12-10 00:12:03.0 +0100 @@ -7,7 +7,11 @@ db_purge rm -rf /var/cache/localepurge rm -f /etc/locale.nopurge -/usr/bin/ucf --purge /etc/locale.nopurge +if [ -x /usr/bin/ucf ]; then +/usr/bin/ucf --purge /etc/locale.nopurge +else +echo 2 /etc/locale.nopurge could not be purged because ucf was not found. +fi rm -f /etc/locale.nopurge.md5sum echo
Bug#424776: achims-guestbook: affected by php4-removal
Hello, Since the 0-day NMU policy is still in force (see [1]), I'm going to do a Non-maintainer upload now to fix this bug and an other policy violation (debhelper must be listed in Build-Depends because it is required for the clean target in debian/rules). The diff between the current version and my NMU is available at: http://people.debian.org/~fst/NMUs/achims-guestbook_2.53-1_2.53-1.1.diff Greetings, Frank [1] http://lists.debian.org/debian-devel-announce/2007/09/msg0.html signature.asc Description: This is a digitally signed message part.
Bug#424776: achims-guestbook: affected by php4-removal
On Friday 02 November 2007, James Westby wrote: php4 is no longer available, so the dependency should just be changed, rather than an alternative added. Quoting http://wiki.debian.org/PHP4Removal: | fix-deps | packages with incorrect dependencies (i.e. depending on only php4 instead of | php5 | php4). Grüße, Frank signature.asc Description: This is a digitally signed message part.
Bug#431516: depends on non-essential package adduser in postrm
package greylistd tags 431516 + patch thanks Hello, One way to fix this bug is to mask the calls to deluser/delgroup with tests that check for the existence of deluser/delgroup as outlined in this wiki page: http://wiki.debian.org/AccountHandlingInMaintainerScripts. If adduser is available it will remove the user/group otherwise a warning is printed that the user/group couldn't be deleted because the deluser/delgroup commands couldn't be found. A patch to fix this is attached. Grüße, Frank diff -Naur greylistd-0.8.3.5.bak/debian/postrm greylistd-0.8.3.5/debian/postrm --- greylistd-0.8.3.5.bak/debian/postrm 2004-12-31 14:54:44.0 +0100 +++ greylistd-0.8.3.5/debian/postrm 2007-08-31 14:17:59.0 +0200 @@ -1,5 +1,5 @@ #! /bin/sh -e -# postinst script for greylistd +# postrm script for greylistd # # see: dh_installdeb(1) @@ -33,8 +33,18 @@ if id -u greylist /dev/null 21 then - deluser --system greylist - delgroup --system greylist /dev/null 21 || true + if [ -x `which deluser 2/dev/null` ]; then + deluser --system greylist + else + echo 2 Not removing \`greylist' system account \ + because deluser command was not found. + fi + if [ -x `which delgroup 2/dev/null` ]; then + delgroup --system greylist /dev/null 21 || true + else + echo 2 Not removing \`greylist' system group \ + because delgroup command was not found. + fi fi ;; signature.asc Description: This is a digitally signed message part.
Bug#431522: depends on non-essential package ucf in postrm
package localepurge tags 431522 + patch thanks Hello, To fix this the ucf call must be masked by a test that checks for the existence of /usr/bin/ucf. The attached patch does exactly this. It additionally prints a warning that ucf could not be found if it is unavailable during purge. Grüße, Frank diff -Naur localepurge-0.5.9.bak/debian/postrm localepurge-0.5.9/debian/postrm --- localepurge-0.5.9.bak/debian/postrm 2007-08-31 17:56:48.0 +0200 +++ localepurge-0.5.9/debian/postrm 2007-08-31 19:10:52.0 +0200 @@ -7,7 +7,11 @@ db_purge rm -rf /var/cache/localepurge rm -f /etc/locale.nopurge -/usr/bin/ucf --purge /etc/locale.nopurge +if [ -x /usr/bin/ucf ]; then +/usr/bin/ucf --purge /etc/locale.nopurge +else +echo 2 /etc/locale.nopurge could not be purged because ucf was not found. +fi rm -f /etc/locale.nopurge.md5sum echo signature.asc Description: This is a digitally signed message part.
Bug#424776: achims-guestbook: affected by php4-removal
Hello, This package's dependencies are currently: Depends: php4, apache | httpd I've downloaded it, added php5 as alternative dependency to php4, built and tested it with PHP 5. It seems that the package works with PHP 5 without any problems. Therefore too fix this bug it is sufficient to add php5 as alternative dpendency, see the attached patch. Grüße, Frank diff -Naur achims-guestbook-2.53.orig/debian/control achims-guestbook-2.53/debian/control --- achims-guestbook-2.53.orig/debian/control 2007-08-30 19:15:06.0 +0200 +++ achims-guestbook-2.53/debian/control 2007-08-29 17:47:29.0 +0200 @@ -7,7 +7,7 @@ Package: achims-guestbook Architecture: all -Depends: php4, apache | httpd +Depends: php5 | php4, apache | httpd Description: php driven guestbook Achims Guestbook (http://www.lkcc.org:8500/) is a php driven guestbook that relies on text files for storing entries. It has support for ip signature.asc Description: This is a digitally signed message part.
Bug#402519: Reopen
Hi Filipus, On Monday 02 July 2007, Filipus Klutiero wrote: Le lundi 2 juillet 2007 06:04, Frank S. Thomas a écrit : Hi Filipus (or Philippe?), On Wednesday 21 March 2007, Filipus Klutiero wrote: reopen 402519 severity 402519 grave thanks [...] I understand that you reopened this bug because it is still found in the current kernel-patch-debianlogo (1.5), but could you please explain why you raised it's severity from important to grave? Yes. The package is now unusable in stable, testing and unstable. (A bug preventing a package from being used in a suite is grave) You mean because there is no kernel source package 2.6.18 in all three suites? Still the diff in /usr/src/kernel-patches/diffs/debianlogo/ can be applied manually, hence I find severity important more appropriate. However, as soon as the new kernel-patch-debianlogo version, which is an empty transitional package for easier upgrades to linux-patch-debianlogo, enters the archive, this bug vanishes and can be closed. Grüße, Frank
Bug#402519: Reopen
Hi Filipus (or Philippe?), On Wednesday 21 March 2007, Filipus Klutiero wrote: reopen 402519 severity 402519 grave thanks [...] I understand that you reopened this bug because it is still found in the current kernel-patch-debianlogo (1.5), but could you please explain why you raised it's severity from important to grave? Since the package is still usable with older kernels, I think severity important is more appropriate. Thanks, Frank -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#430285: ldbl128 transition for alpha, powerpc, sparc, s390
Hi, On Saturday 23 June 2007, Matthias Klose wrote: Package: liborsa0-dev [...] This package has been indentified as one with header files in /usr/include matching 'long *double'. Please close this bug report if it is a false positive, or rename the package accordingly. I think this is a false positive, the only occurrence of long double is in a macro definition: $ grep -r long.*double * orsa_common.h:#define PI 3.1415926535897932384626433832795029L /* long double constant, thanks to the trailing L */ And this constant is only used as follows: const double halfpi = PI/2; const double pi = PI; const double twopi = PI+PI; const double pisq = PI*PI; Could you please confirm that the package does not need to be renamed? Thanks, Frank signature.asc Description: This is a digitally signed message part.
Bug#422643: FTBFS: /usr/share/boinc-dev/lib/app_ipc.h:30:25: error: common_defs.h: No such file or directory
package boinc-app-seti tags 422643 + confirmed pending stop On Monday 07 May 2007 15:27, Michael Ablassmeier wrote: Lucas has rebuild the archive on i386 and your package Failed to Build from Source with the following error: de/openssl -c -o seti_boinc-main.o `test -f 'main.cpp' || echo './'`main.cpp In file included from /usr/share/boinc-dev/api/boinc_api.h:107, from main.cpp:59: /usr/share/boinc-dev/lib/app_ipc.h:30:25: error: common_defs.h: No such file or directory /usr/share/boinc-dev/lib/app_ipc.h:233: error: 'NGRAPHICS_MSGS' was not declared in this scope make[3]: *** [seti_boinc-main.o] Error 1 This is due the missing header common_defs.h in the boinc-dev package. I've already added it to boinc-dev in our Subversion repository so boinc-app-seti will be buildable again after the next boinc upload. Grüße, Frank pgpceCLiHC5I9.pgp Description: PGP signature
Bug#422528: boinc-app-seti: x86_64 package does not work: seti says no such platform
Hi, On Sunday 06 May 2007 20:22, Folkert van Heusden wrote: 2007-05-06 11:40:21 [EMAIL PROTECTED] Message from server: platform 'x86_64-pc-linux-gnu' not found Could you please run update-boinc-applinks --create as root, restart the BOINC client and check if the client uses the anonymous platform for [EMAIL PROTECTED] then? If you don't use Debian's default data directory /var/lib/boinc-client, you need to run: update-boinc-applinks --create --data-dir=/path/to/your/BOINC-data-dir Is there anywhere in the logs a hint that the anonymous platform mechanism for [EMAIL PROTECTED] is used? pgpXZ0mTJ9Glq.pgp Description: PGP signature
Bug#379137: SIGSEGV: segmentation violation
clone 379137 -1 notfound 379137 5.4.11-4 retitle -1 boinc-client segfaults on startup submitter -1 Thomas K. [EMAIL PROTECTED] severity -1 important found -1 5.4.11-3 found -1 5.4.11-4 thanks Since the bug Thomas reported is different from the original bug that was reported as #379137, I'm splitting his report off into its own bug. - Frank pgp4Lm6sgB11h.pgp Description: PGP signature
Bug#379137: SIGSEGV: segmentation violation
Hi, On Tuesday 19 December 2006 17:43, Thomas K wrote: I did not attach the output of strace boinc_client because the file was about 40 Mb big. But by looking at that output I figured something out: If I create a new directory and move all the boinc-xml-files (client_state.xml, client_state_prev.xml and so on) to it and then run boinc_client from that directory it does not segfault. For me it seems as boinc_client is parsing every subdirectory of the directory it was startetd from (every subdirectory appeared in the output of strace). Before I always started it from my home so there may be just too many directories for it. But I guess you will know better than me about that ;) Without looking at the code I'd think that the BOINC client only parses the projects/ and slots/ directory in the current working directory, but I'm not sure if it even does this. If you start the client, that you got from the BOINC website, in your home directory does it also segfault? And could you please test if version 5.4.11-3 also segfaults? You can download it here: http://ftp.debian.org/debian/pool/main/b/boinc/boinc-client_5.4.11-3_i386.deb I'm still interested in the strace output, the last 100 lines are enough: tail -n 100 strace_output_file And it would be nice if you could run the BOINC client in the GNU Debugger (gdb) and send me the output, because I can't reproduce this bug, and therefore I can't debug it. Grüße, Frank P.S.: Please also CC [EMAIL PROTECTED] when replying, thanks. -- Anmut bringen wir ins Leben; Leget Anmut in das Geben. pgpwgJq0LISlL.pgp Description: PGP signature
Bug#379137: SIGSEGV: segmentation violation
Hi, On Friday 15 December 2006 01:14, Thomas K. wrote: Package: boinc-client Version: 5.4.11-4 Followup-For: Bug #379137 boinc_client gets segfault on launch. version from boinc-website works fine. [...] 2006-12-15 01:04:50 [---] No general preferences found - using BOINC defaults 2006-12-15 01:04:50 [---] Local control only allowed 2006-12-15 01:04:50 [---] Listening on port 31416 SIGSEGV: segmentation violationSpeicherzugriffsfehler Could you please install strace: apt-get install strace And then run strace boinc_client and send me the output of this command. Make sure that no other BOINC client is running by calling: /etc/init.d/boinc-client stop and killall boinc_client Grüße, Frank pgpZBzkduXjrh.pgp Description: PGP signature
Bug#401600: constructor/de-constructor never found for all testprograms build against boinc-dev
package boinc-dev
severity 401600 normal
thanks
On Monday 04 December 2006 20:53, Folkert van Heusden wrote:
Package: boinc-dev
Version: 5.4.11-3
Severity: grave
Justification: renders package unusable
Nothing builds with this library.
That's not true. The boinc-app-seti package builds with this package
(therefore I'm downgrading the severity of this bug). Maybe you should have a
look at boinc-app-seti's makefiles to figure out how to use BOINC libraries
and headers.
For example:
---
#include stdio.h
#include BOINC/boinc_api.h
int main(int argc, char *argv[])
{
APP_INIT_DATA aid;
return 0;
}
Compile with: g++ -lstdc++ -lboinc_api -lboinc test.cpp
gives:
/tmp/ccxYULhN.o: In function `main':
test.cpp:(.text+0x1f): undefined reference to
`APP_INIT_DATA::APP_INIT_DATA()' test.cpp:(.text+0x32): undefined reference
to `APP_INIT_DATA::~APP_INIT_DATA()' collect2: ld returned 1 exit status
Other example:
-
#include BOINC/lib/gui_rpc_client.h
int main(int argc, char *argv[])
{
RPC_CLIENT test;
return 0;
}
Compile with: g++ -I /usr/include/BOINC/ -lstdc++ -lboinc_api -lboinc
rpctest.cpp gives:
/tmp/ccfemaKV.o: In function `main':
rpctest.cpp:(.text+0x19): undefined reference to `RPC_CLIENT::RPC_CLIENT()'
rpctest.cpp:(.text+0x29): undefined reference to
`RPC_CLIENT::~RPC_CLIENT()' collect2: ld returned 1 exit status
pgppje2iY6B6e.pgp
Description: PGP signature
Bug#379137: boinc-client errors out with a Floating point exception error when application is launched.
On Friday 21 July 2006 18:09, Jim Woodruff wrote: Subject: boinc-client errors out with a Floating point exception error when application is launched. I think the CFLAGS option -ffast-math is the culprit. Could you compile boinc without -ffast-math and try out if the BOINC core client still throws the floating point exception? - Frank pgpNWRzbmdm31.pgp Description: PGP signature
Bug#369288: boinc-client: doesn't start and gives SIGSEGV
severity 369288 important tags 369288 + unreproducible moreinfo stop Hi, On Sunday 28 May 2006 23:01, aldo wrote: starting from today, when i run boinc, it starts and tries to resume a task but immediately exits giving the followinf error (repeated several times): SIGSEGV: segmentation violation Stack trace (64 frames): boinc[0x8089fab] [0xe420] /lib/libgcc_s.so.1[0xb79ac7ab] /lib/libgcc_s.so.1(_Unwind_Backtrace+0x3c)[0xb79ad3cc] /lib/tls/i686/cmov/libc.so.6(backtrace+0x7c)[0xb7a9814c] I can't reproduce this on my system, so I need more info from you. To what projects are you attached? To what project does the task belong that the BOINC client is trying to resume? Did you upgraded your system yesterday? Could you please shutdown all running BOINC clients (killall boinc killall boinc_client) and start boinc in an empty directory. Does is start then without crashing? Grüße, Frank -- Die Garde stirbt, aber sie ergibt sich nicht! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#340590: libginac1.3c2a depends on nonexistent libcln4 on i386
Package: libginac1.3c2a Severity: grave Justification: renders package unusable Hi, libginac1.3c2a is not installable on i386 because it depends on libcln4 which is not available in unstable: apt-get install libginac1.3c2a [...] The following packages have unmet dependencies: libginac1.3c2a: Depends: libcln4 but it is not installable E: Broken packages - Frank -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (990, 'unstable'), (1, 'experimental') Architecture: i386 (i686) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.14 Locale: LANG=C, LC_CTYPE=de_DE.UTF-8 (charmap=UTF-8) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#339246: library package needs to be renamed (libstdc++ allocator change)
block 339246 by 339172 thanks On Thursday 17 November 2005 03:22, Matthias Klose wrote: Compiler versions g++-4.0_4.0.2-4 and g++-3.4_3.4.4-10 are now in the archive. The renaming of the library packages can now start. You can upload the packages even before the toolchain is built for all architectures because the packages with the new binary packages will be hold in the NEW queue until the required toolchain changes are installed on the buildd's. Just for the records: orsa depends on the ginac library which also has to be renamed (libginac1.3c2 - libginac1.3c2a), hence I'll ask my sponsor for an upload after the renamed ginac package is available on all architectures. - Frank -- The best way to accelerate any computer running Windows is at 9,81 m/s^2. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
